text
stringlengths 8
5.74M
| label
stringclasses 3
values | educational_prob
sequencelengths 3
3
|
|---|---|---|
1. Field of the Invention The present invention relates to airfoils used in a gas turbine airfoil, and more specifically to an airfoil having a platform. 2. Description of the Related Art Including Information Disclosed Under 37 CFR 1.97 and 1.98 In a gas turbine engine, a turbine section includes a plurality of turbine blades and guide nozzles or vanes on which a hot gas stream reacts to drive the turbine. This hot gas stream passes through and around the turbine blades. A hot gas migration phenomenon on the airfoil pressure side is created by a combination of hot flow core axial velocity and static pressure gradient exerting on the surfaces of the airfoil pressure wall and the suction wall of adjacent airfoils. As a result of this hot gas flow phenomena, some of the hot core gas flow from an upper airfoil span is transferred toward a close proximity of the platform and therefore creates a high heat transfer coefficient and high gas temperature region at approximately two-thirds of the blade chord location. FIG. 1 shows a cut-away view of the vortices formation for the hot glow gas migration across the turbine flow passage, and shows a hot spot on the platform of each blade on the pressure side 18. A Prior Art blade with platform is shown in FIG. 2. The blade includes a root 10, a cooling fluid passage 12, a platform 14, an airfoil 18, and a tip 19. A fillet region 16 is formed between the airfoil and the platform. Cooling of a blade fillet region 16 by means of conventional backside convective cooling method yields inefficient results due to the thickness of the airfoil fillet region 16. On the other hand, drilling film holes 17 at the blade fillet to provide for film cooling produces unacceptable stress by the film cooling holes 17. A line of film cooling holes 17 along the lower section of the blade for cooling the blade fillet region 16 would be located in the region of the airfoil having the highest pull stress levels, thereby providing a point of weakness at the highest stress points on the blade. U.S. Pat. No. 6,341,939 B1 issued to Lee on Jan. 29, 2002 entitled TANDEM COOLING TURBINE BLADE discloses a turbine blade with a central cooling air passage and a metering hole leading from the central passage and onto the outer surface of the platform around the transition region of the blade for cooling the transition region (space between the airfoil and the platform). However, the Lee invention does not uncouple the airfoil from the platform as does the present invention, among other differences. U.S. Pat. No. 5,340,278 issued to Magowan on Aug. 23, 1994 entitled ROTOR BLADE WITH INTEGRAL PLATFORM AND A FILLET COOLING PASSAGE discloses a turbine blade with a cooling fluid passage connecting the core passage of the blade with the damper or dead rim cavity for the purpose of cooling the fillet of the platform and airfoil transition. No cooling air passes onto the outer surface of the airfoil platform or airfoil, and the platform is not uncoupled from the airfoil as in the present invention, among other differences. U.S. Pat. No. 5,382,135 issued to Green on Jan. 17, 1995 entitled ROTOR BLADE WITH COOLED INTEGRAL PLATFORM shows a turbine blade with a platform having a plurality of cooling holes located on the pressure side of the blade for cooling the platform. A row of cooling holes closest to the airfoil surface are supplied with cooling air from the core or central passage of the blade, while an outer row of cooling holes are supplied with cooling air from the dead rim cavity below the platform. The Green invention does not provide for the uncoupling of the platform from the airfoil as in the present invention, among other differences. One alternate way of cooling the fillet region is by the injection of film cooling air at discrete locations along the airfoil peripheral into the downward hot gas flow to create a film cooling layer for the fillet region 16. However, in order to achieve a high film effectiveness level, the discrete holes used in this type of film cooling injection have to be in a close pack formation. Otherwise, the spacing between the discrete film cooling holes and areas immediately downstream of the spacing are exposed to less cooling or no film cooling air at all. Consequently, these areas are more susceptible to thermal degradation and over temperature. On the other hand, the close pack cooling holes at the blade lower span becomes undesirable and the stress rupture capability of the blade is lower. An object of the present invention is to reduce or eliminate the high heat transfer coefficient and high gas temperature region as well as high thermal gradient problem associated with a turbine blade platform. Another object of the present invention is to uncouple the platform from the airfoil of the blade in order to reduce stress from thermal gradients between the two parts of the blade. | Low | [
0.5106837606837601,
29.875,
28.625
] |
WEBVTT Tonight police are investigating an ATM burglary that happened at a Des Moines convenient store early this morning. Just after 3 a.m., a suspect broke into the Valero convenient store at 4140 southwest Park Avenue through the rear door. Police say the suspect cut wires disabling security alarms. He even crawled across floor to avoid being seen and broke into the ATM at the front of store. But security cameras caught it all ... That is what you are seeing right now. The suspect dressed in all black wearing gloves and a ski mask. The store manager says the suspect stole about 800-dollars. < " VERY DISCOURAGING. PLUS IT MAKES MY CASHIERS NERVOUS, WHICH I CAN'T BLAME THEM, AND OF COURSE I DON'T WANT ANY OF THEM TO BE NERVOUS, WHILE THEY WORK HERE." The suspect also stole a Gatorade drink during the burglary. Police are still investigating. | Low | [
0.503703703703703,
34,
33.5
] |
Rack toothing for linear drive (versions) SUBSTANCE: invention relates to tooth rack gear, transforming rotary motion into forward motion and on the contrary. It can be used instead of usual involute gear of rack-and-gear mechanisms in linear drives of stands, in devices of auto steering control, and also in load-lifting technology (rack-and-gear jack and suchlike). Rack toothing for linear drive contains wheel (1) and rack (2). Wheel is implemented with one tooth with profile in transverse section of wheel in the form of circumference (3), off-centre displaced from pivot pin of wheel (1). Skew tooth is formed by successive and continuous turn of transverse sections around wheel's axis with formation of helical surface. Rack (2) allows helical teeth (4) of cycloidal profile, conjugated to helical surface of wheel (1). In the second option wheel (1) allows also one tooth and composed from certain crowns, each of which allows profile in the form of off-centre shifted circumference. Teeth of rack (2) are composed also from several cycloidal crowns. Adjacent wheel crowns and racks are phase shifted relative to each other for pitch of corresponding crown, divided for number of crowns. The invention relates to timing of kinematic pairs, and more particularly to a rack and pinion transmission that converts rotary motion to linear motion and Vice versa. It can be used instead of the usual rack and pinion mechanisms in linear actuators, machines, devices steering of vehicles and lifting equipment (rack and pinion jacks and the like). Known cylindrical gear with a rack and pinion gearing, in which a gear wheel is in mesh with the rack, forming a fixed part of the progressive pair (Afraide. Dictionary-reference mechanisms. M: "engineering", 1987, s). As engagement in pinion used involute toothing. Axle gear perpendicular to the direction of movement of the rack. The main disadvantage of the rack and pinion is its low load capacity. To zoom, you need to increase the modulus of the tooth, and hence the size of gears. This increases the linear speed of movement of the rack at the same speed of rotation of the wheel, which is not always valid. Known rack and pinion worm gear (see ibid., str). It is similar to the conventional worm gear, in which the worm wheel is replaced by rail. The axis of the worm is usually with the direction of movement of the rack some angle, which is dependent upon the angle of helix of the worm and the angle of the teeth of the rack. The lack of a rack and pinion worm gear is the same as a conventional worm gear, namely a low efficiency, not exceeding in most designs of 0.5. A known mechanism for converting rotary motion into reciprocating (US 5187994) and later modifications of this mechanism (EP 0770795, US 5477741, US 5699604), which in the most General case consists of two parallel shaft rotated synchronously from one drive, a few eccentrics mounted on each shaft with a phase difference, as well as plates with teeth and holes, which are rotatably installed these eccentrics. The mechanism also includes a rack, which interact with the teeth of the plates. When the rotation shaft plate commit plane of planetary motion and, siteplease with the teeth of the rail, move it in the longitudinal direction. The profile of the teeth of the plates in one embodiment, has a semicircular shape, and the teeth of the slats have a cycloidal profile. In the second variant, the teeth of the slats are of semicircular profile and the teeth of the plates of a cycloidal profile. To reduce friction teeth semicircular profile are in the form of rollers freely rotatable on axes. The mechanism of such engagement can be quite high efficiency at high load capacity, due to the fact that motion is transmitted several the parallel branches. The main disadvantage of this device is its complexity due to the large number of parts. In addition, the device is critical to lack of synchronism of rotation of the shafts and require high precision Assembly. For the prototype will choose the conventional rack and pinion engages the teeth of involute profile described above (Afraide. Dictionary-reference mechanisms. M: "engineering", 1987. s), as having the largest total number of the proposed solution characteristics. It contains being in engagement gear and the rack. The wheel and the rail is installed with the possibility of translational movement relative to each other. The speed of the translational movement of the slats is directly proportional to the diameter of the wheel. The load capacity of the gear depends on the module of the teeth, and hence the diameter of the wheel. Increase the load capacity leads to an increase in the diameter of the wheel and the speed of movement of the rack when other conditions are equal. Thus, the object of the invention is to provide a compact, simple and reliable rack and pinion gearing for linear actuator. The technical result achieved by the invention is to increase the load capacity of gear with the same dimensions, and the possibility of obtaining high speeds Reiki regardless of the size of the wheel, (and depending who it only from the corner of the step slats). To solve this problem a rack and pinion gearing, as the prototype comprises a gear wheel and the rack installed with the possibility of translational movement relative to each other. Unlike the prototype of the wheel and the rail is made with oblique teeth, and the wheel is made with a single tooth profile in the mechanical section of the wheel in a circle eccentric offset from the axis of rotation of the wheel, and oblique tooth wheel is formed of a consistent and continuous rotation of these end sections around the axis of the wheel with the formation of the helical surface, and the rail has helical teeth of a cycloidal profile associated with the helical surface of the wheel and providing a linear contact of the teeth. For continuous transmission of the movement angle of the axial overlap gears and Reiki to exceed 180 degrees. Like any helical gearing, the gearing will have axial components of the force. In order to eliminate them, the teeth of the wheel and the rail is advisable to make herringbone. Possible and the second embodiment of the inventive concept, forming the essence of the present invention. The toothed profiles of the wheels and rails in this case are not helical, and a compound of the packages, at least three identical crowns. Crowns in each packet is shifted in phase to each otnositelnogo a step corresponding to the crown, divided by the number of crowns. Each crown wheel is an eccentric offset from the axis of the wheel circumference, and each crown Reiki has the teeth of a cycloidal profile. As the wheel in the proposed mesh has only one tooth, individual crowns must be rotated relative to each other by an angle equal to 360 degrees divided by the number of crowns. The wheel is composed of three crowns, crowns rotated relative to each other at an angle of 120 degrees. For Reiki with three crowns crowns are shifted relative to each other along the direction of the translational movement of the rack 1/3 of the linear step rail. Increasing the number of crowns with a simultaneous decrease in the thickness of the second variant of the invention seeks to the first. The invention is illustrated in graphic materials, in which figure 1 is a perspective view of the first variant of the proposed rack and pinion gearing. Figure 2 shows the same gearing, end view. Figure 3 and 4 shows diagrams illustrating the formation of cycloidal profile rack with different linear step. Figure 5 illustrates the engagement of the helical teeth. Figure 6 shows a General view of the second variant of the gear wheel and rack, composed of 6 crowns. The mesh depicted in figure 1 and 2, is formed by the wheel 1 and the rack 2. The axis of rotation OO1wheel 1 perp is ndikumana the direction of translational movement of the slats 2, shown in the figures by the arrow. The wheel 1 is made with one helical tooth angle of the axial overlap in 360 degrees. Helical tooth formed a consistent and continuous rotation for each end section of the wheel 1, which represents the eccentric offset by the distance e from the axis of rotation of the wheel circumference 3. Or the formation of helical tooth can be considered as uniform and continuous movement of the eccentric offset of the circle 3 along the axis OO1with simultaneous continuous rotation of this circle around the axis OO1. In the figures, the dashed lines 3', 3”, 3”',... marked circle in the end sections of the wheel 1, is made through 60 degrees. Helical teeth 4 of the rail 2 in the end sections have the form of a cycloidal curve 5. Figures 5, 5', 5”, 5”' marked cycloidal curves in adjacent sections of the rail, the mating with the respective eccentric circles 3, 3', 3”, 3”' in the end sections of the wheel 1. Because the construction of each of the circles 3, 3', 3”, 3'”,..., has a point contact with the corresponding cycloidal curves 5, 5', 5”, 5”',..., the toothed wheel 1 with the rail 2 will have a continuous line of contact of the aircraft. To build paired with a wheel 1 of a cycloidal profile 5 rack 2 refer to the diagrams in figure 3 and 4. The letter marked On the center of rotation of the wheels is 1. Its jagged profile here represented by circle 3, the center of which is offset by the distance e from the center of rotation O. To obtain a cycloidal profile 5 of the rail 2 will build from the center Of the circle 6 on circle 3. Rolling this circle 6 direct 7 without slipping her point And coincident with the center of the circle 3 and located within the circle 6, describes trochoid (shortened cycloid) 8 increments of t6. Equidistant this trochoid 8 shifted by a distance r (here r is the radius of the circle 3 in the end section of the wheel 1), forms the desired cycloidal curve 5. For more steep fronts cycloidal teeth Reiki 2 you should build trochoid in smaller increments. It is formed by rolling circumference 9 of smaller diameter, as shown in figure 4. Circle 9 there is also built with a center point On, but its diameter D9less than the diameter DGforming a circle 6 figure 3. Rolling without slipping circumference 9 straight 10 point And inside the circle 9, describe trochoid (shortened cycloid) 11, which has more sharp edges than the curve 8. Equidistant trochoid 11 and will be required cycloidal curve 12, which forms the profile of the tooth 4 of the rail 2 in smaller increments t12. For wheel 1 having a cross-section circle of the same radius r with one and the e the eccentricity e, you can build a lot of cycloidal curves, characterized by the step size, and hence the steepness of the front teeth. The reduction step of the tooth Reiki 2 is limited only by the effect of cutting teeth. Obviously, in the engaged figure 3 path length DE, passable at any point on the circumference 3 for one full turnover, approximately equal to the length of this circle. Engaged as in figure 4, the circumference 3 is significantly greater than the corresponding length of the path of GH. Therefore, the mesh in figure 3 will have less slippage than engagement with the greater curvature of the front teeth of the rail 4. However, this engagement is significantly worse than the existing distribution of power. Indeed, the interaction force F of the tooth of the wheel 1 and the tooth Reiki 2 is directed perpendicular to the profiles of the teeth at the point of contact. This force can be resolved into two mutually perpendicular components of F1and F2. Component F1directed along the rail and is actually the driving force and the component of F2is the force with which the wheel 1 is pressed to the river 2 (pressure force). Obviously, with equal moments of rotation of the wheel 1 operating force F1for Reiki in smaller increments will be greater than the operating force for engagement of figure 3. Thus, in each case for the same wheel 1 step selection Reiki 2 BU the et requirements for power or efficiency. As everyone helical gearing, the proposed mesh has a component of force directed along the axis of the wheel and tending to shift the wheel relative to the rail. To prevent axial displacement of the rack and pinion transmission, you should use angular contact bearings. Another solution to this problem is the implementation of the rims of the wheels and rails Chevron, as shown in figure 5. The wheel 1 has the length of two sections 13 and 14 formed helical surfaces opposite direction. Circle 3 in the end section of the wheel on the section 13 has a continuous rotation around the eccentric offset of the axis OO1clockwise, and on plot 14 counterclockwise. Similarly, gear Reiki 2 consists of two sections 15 and with the right left 16 cycloidal teeth formed phase shift of a cycloidal curve 5 in opposite directions. It is obvious that due to the symmetry of the arrangement of the teeth of the axial components of force in a Chevron engaged mutually balanced. With all the advantages of the proposed rack and pinion gearing is quite difficult to manufacture, requires a multi-axis CNC machines. This idea of engagement can be implemented in another embodiment, more simple to manufacture. Refer to the diagram of the education wintopo what about the profile of the tooth of the wheel 1, depicted in figure 1 and 2. If the profile of the tooth of the wheel 1 to obtain continuous rotation and offset of the eccentric circle 3 relative to the axis of rotation OO1wheel 1, and to separate these two movements, you will receive a stepped profile of the wheel 1, an educated individual rotated relative to each other the same rims 17, 17', 17”, ... (see Fig.6). Every crown 17 is formed by a cylinder with an eccentric offset circle in cross section. Neighboring crowns 17, 17' are rotated relative to each other by an angle equal to the angular step of the wheel divided by the number of crowns. Figure 6 angular step of the wheel 1 is 360 degrees, the number of crowns is 6. Therefore, neighboring crowns 17 will be rotated relative to each other by 60 degrees. To produce such a stepped profile wheels can be either single crowns, rigidly fastened together, or by the wheel with stepped profile in the form of single parts like the crankshaft. Similarly constructed and compound gear profile Reiki 2, only individual crowns 18, 18', 18”, 18”', ..., shifted relative to each other along the rails by a distance equal to the step slats, divided by the number of crowns. In the General case about the crowns of the composite wheel and compound Reiki can say that they are displaced relative to each other in phase and the offset is step corresponding crown, divided by the number of vents is C. Each pair of the rims 17 and 18 wheels 1 and Reiki 2 contacts in a straight line and a common line contact profile is a piecewise continuous polyline curve. In engagement with a stepped profile wheels and rails no problem the axial component of the force, as it can be seen as a superposition of pairwise links separate spur crowns. It should be noted that by increasing the number of crowns in engagement, we'll come to the first embodiment of the gearing of helical screw teeth. In turn, mesh with the helical teeth can be considered as the engagement of the stepped profiles, where the number of crowns infinitely large, and the displacement in phase between adjacent crowns infinitely small. Consider the operation of the gearing shown in Fig.1-2. During the rotation of the wheel 1 about the axis OO1eccentrically located relative to the axis of the circle 3 (3', 3”, 3”' and so on) in any of the mechanical section of the wheel 1 is in contact with a cycloidal profile 5 (5', 5”, 5”', ...) Reiki 2 in the same section. Let the wheel 1 rotates clockwise, as shown in the figures. Circle 3 in the frontal plane of the mesh (see figure 2), in contact with the top of the cycloidal tooth 5, when rotating around the center O begins to put pressure on the tooth, causing the movement of rail 2 in the same direction by an amount equal to half of its angular step. After the floor is guilty of revolution of the wheel 1 circle 3 will come into contact with the depression of a cycloidal profile 5 of the rail 2 and on the next half of the turnover in this section forces acting on the rail 2 will not. Similar reasoning can also lead to other mechanical sections of the wheel 1 and Reiki 2, where the power contact siteplease profiles will be carried out only on the half turn of the wheel 1. If the angle of the axial overlap of the wheel 1 will be equal to or greater than 180 degrees, the power contact is set to the full revolution of the wheel 1. This means that the movement of rail 2 is continuous and one revolution of the wheel 1 rail 2 will be moved in the longitudinal direction on one tooth. I.e. movement speed rail 2 is determined only step her teeth and does not depend on the diameter of the wheel 1, as it was in the prototype. In addition, under the same diameters of the wheels in the prototype and in the present invention the dimensions of the tooth wheel and rail according to the invention will be significantly larger than that of the prototype with involute gearing. Thus, significantly higher and the load capacity of the proposed rack and pinion gearing. If the angle of the axial overlap of the wheel 1 is less than 180 degrees, then the mesh will appear "dead zones", where the motion is not passed. However, the engagement and in this case, it may be serviceable if the desired reciprocating movement of the slats 2 smaller step angle. In addition, gearing operable and when the wheel 1 has a large centrifugal mass, the inertia of which Preodolenie the "dead zone" of the engagement. Work rack and pinion type gearing according to the second variant does not differ from the above. The motion passed successively different pairs of crowns composite wheels and composite slats, i.e. crowns 17-18, 17'-18', 17”-18” etc. 1. Rack and pinion gearing for linear drive with toothed wheel and the rack installed with the possibility of translational movement relative to each other and are in engagement, wherein the wheel and the rail is made with oblique teeth, and the wheel is made with a single tooth profile in the mechanical section of the wheel in a circle eccentric offset from the axis of rotation of the wheel, and oblique tooth wheel is formed of a consistent and continuous rotation of these end sections around the axis of the wheel with the formation of the helical surface, and the rail has helical teeth of a cycloidal profile associated with the helical surface of the wheel and providing a linear contact of the teeth. 2. Rack and pinion type gearing according to claim 1, characterized in that the angle of the axial overlap of the wheel exceeds 180°. 3. Rack and pinion type gearing according to claim 1 or 2, characterized in that the teeth of the wheels and rails made herringbone. 4. Rack and pinion gearing for linear drive with toothed wheel and the rack installed with the possibility of translational movement relative to the part of each other and are in engagement, characterized in that the toothed profile and wheels, and Reiki is made in the form of a package of at least three identical toothed crowns each, crowns in each packet are shifted relative to each other in phase by step the corresponding crown, divided by the number of crowns, and tightly connected to each other, each gear wheel has a tooth profile in which the mechanical section defined eccentrically offset from the axis of the wheel circumference, and each gear Reiki has the profile of a cycloidal shape. SUBSTANCE: invention is related to steering mechanisms of wheel transport and traction machines, and may also find application in motor industry, tractor construction, machine-tool construction, instrument making, production of weight-lifting machines, where mechanisms are required with preset patterns of their separate work members movement. Steering mechanism comprises tooth gear with master gear (1) and tooth rack (6). Tooth rack (6) is made in the form of two parallel parts that move one along each other. Teeth (7) of racks are installed with the possibility of alternate engagement with gear in radiuses of its dividing circumferences, which makes it possible to obtain an alternation transmission ratio. EFFECT: makes it possible to improve controllability of transport vehicle due to provision of optimal pattern for variation of transmission ratio of rack steering mechanism, and also to increase traffic safety. SUBSTANCE: invention can be used in systems with piston, slider and other units. Proposed mechanism has housing, guides 2, slider 3, output shaft 4 with circumferential section 5 of gear rim alternately engaging with two sections 6, 7 of rim (racks) arranged from side of slider 3 with teeth orientated in opposite direction. End working teeth of edges of rim sections 5-7 are made with body truncation surface having at least one bend and orientated with tilting at acute angle by one extremity to surface of point and by other extremity, square to one of end faces. Surface of truncation of tooth body together with cut of top surface on opposite sections 5 and 6 or 5 and 7 of rim is orientated so that corresponding pairs of working teeth are aligned with ridges at insignificant clearance at opposite relative displacement. SUBSTANCE: invention relates to lever-tooth mechanisms with rack and it can be used to drive mechanisms of slotting machine, automobile windshield wiper, brush cleaner of screens of grain-cleaning machines, in automatic machines, etc. Proposed mechanism contains base, toothed wheel and crank hinge-mounted on base for rotation, and toothed rack engaging with toothed wheel. Lever with guide member is hinge-mounted on axle of toothed wheel. Guide member engages with rear side of toothed rack forming translational pair. Guide member is made in form of roller hinge-mounted for rotation on axle installed on free end of lever provided with thread. Axle of roller is provided with through cross hole to pass threaded end of lever. Roller is held on lever by nuts arranged from outer and inner sides of roller to adjust and fix radial clearance in meshing between rack and toothed wheels. SUBSTANCE: motion converter comprises four cylinders grouped in pairs. Pistons of opposite cylinders are connected with each other through fixed bar. The converter has the first and the second pair of fixedly connected parallel toothed racks brought into engagement with corresponding segmental gear-wheels fixedly connected to one output shaft. In each segmented gear-wheel teeth extend for less than half of 180° pitch circle. Segmented gear-wheels cooperating with corresponding toothed racks are shifted through predetermined angle, preferably 90° angle, one relatively another. FIELD: the invention refers to the field of machine building namely to transformers of alternate/reciprocal motion into rotary and vice-versa and may be used in piston engines, pumps, compressors and other gears. SUBSTANCE: the transformer is a planetary toothed mechanism consisting of a non-rotating central pinion with inner toothed gearing making alternate/reciprocal motion relatively to the axle rotating together with firmly connected with it from one side and interacting with the central pinion-satellite. The length of the toothed rim of the satellite is two times smaller then the length of the toothed rim of the central pinion, of the link or of the mechanism providing continuous engagement of the teeth of the central pinion and the satellite. The link and the mechanism are fulfilled in such a manner that install marginally allowable distance between the positions of their joining with the central pinion and with the satellite not installing the maximum allowable value of this distance. The central pinion and the satellite have figured different profile according to their toothed rims. EFFECT: increases the coefficient of efficiency of transformation of the alternate/reciprocal motion into rotary and vice-versa and simplifies the construction of the toothed transformer. SUBSTANCE: according to invention, body 10 forms prismatic chamber 12 whose cross section is oval of odd order formed from arcs 34, 36, 38 with first smaller radius of curvature and arcs 40, 42, 44 with second, larger radius of curvature changing continuously and differentially one into the other. Thus, corresponding cylindrical parts of inner surface of chamber are formed. Chamber 12 accommodates rotating piston 60 whose cross section forms oval of the order smaller by 1 than order of chamber 12. Opposite parts of side surface are formed on rotating piston 60, one of which rotates in part of inner surface of radius of curvature equal to said part and the other adjoins opposite part of inner surface to slide along surface. rotating piston 60 divides chamber 12 in any position into two working spaces 78, 80. Instantaneous axes of rotation 112, 114 of rotating piston 60 are determined on middle plane of piston being fixed for a short time. Working agent to set rotating piston 60 into motion is periodically introduced into working spaces. Rotating piston 60 rotates in each phase of its motion in one of opposite parts of its side surface 70 in corresponding part of inner side surface 62 of chamber around corresponding instantaneous axis of rotation 112 and slides by opposite part of its surface 72 along corresponding opposite part of inner side surface 54 of chamber 12 to stop, i.e. until it comes into extreme position. Then, to execute following phase of movement, instantaneous axis of rotation jumps from previous position into second possible position 114 relative to piston and is fixed in this position for a short rime. Driven or driving shaft 102 is in engagement with rotating piston 60. To prevent kinematic ambiguity of instantaneous axis of rotation in extreme position, instantaneous axis of rotation is mechanically fixed in each extreme position for a time (Fig.1). SUBSTANCE: gearing device comprises two gear racks and gear wheel. The recess is provided between the racks (26) and (28) to prevent gear wheel (10) against the contact with both of the racks simultaneously. The adapting gear member between two racks is defined by the convex gear arc. When gear wheel (10) engages adapting member (30), it disengages rack (26) immediately before the moment when it engages rack (28). SUBSTANCE: gear converter comprises nonrotating shaped gear wheel (1) with inner teeth engagement, which reciprocates with respect to semiaxle (3) that rotates together with shaped pinion (2) secured to the semiaxle from one side and cooperating with shaped gear wheel (1). The length of the gear rim of pinion (2) is twice as less as that of gear rim of shaped gear wheel (1). The continuous engagement of the teeth of shaped gear wheel (1) with pinion (2) is provided by means of the cooperation of cam (4) that is connected with pinion (2) with roller (8) freely rotating around its axis (9). SUBSTANCE: converter comprises the rack, gearing, output shaft, two overloading clutches, and thrusts. The rack connected with the pivot is in coincidence with the overrunning clutches which are in a staggered contact with conical gears. The movement of the conical pivots are limited by means of the pinion secured to the output shaft. The racks provided with the gearing and overrunning clutches are interposed between the two conical gearings. SUBSTANCE: device comprises rack, gears, output shaft, two overrunning clutches, and stops. The output shaft is inclined at an angle of 90° to the driving shaft. The gear engaging the rack is pivotally connected with the driving shaft. The faces of the gear abut against two overrunning clutches which are in a contact with the stop and conical gear pivotally mounted on the driving shaft. The conical gear is connected with the intermediate pinion secured to the output shaft. The intermediate pinion engages the conical gear secured to the driving shaft. The additional stop is secured to the shaft, set in the gear actuated by the rack, and housed by the flange. The device also has roller for keeping the space in the rack-gear engagement constant. SUBSTANCE: gearing converter comprises single gearing frame with inner or outer gear cut, two gearing racks (5) and (16) whose ends are interconnected via intermediate sections that define smooth members for power pinion (13) from one gear rack to the other, locking roller (14) that rolls over the shaped surface, one driving rod (1), carrier (15) with locking roller (14) connected with the gear rack and mounted on power shaft (4), and power pinion (13). The intermediate sections of the gear frame are provided with rounds that are redial with respect to the axis of rotation of the gear rack. SUBSTANCE: device comprises rack, gears, output shaft, two overrunning clutches, and stops. The output shaft is inclined at an angle of 90° to the driving shaft. The gear engaging the rack is pivotally connected with the driving shaft. The faces of the gear abut against two overrunning clutches which are in a contact with the stop and conical gear pivotally mounted on the driving shaft. The conical gear is connected with the intermediate pinion secured to the output shaft. The intermediate pinion engages the conical gear secured to the driving shaft. The additional stop is secured to the shaft, set in the gear actuated by the rack, and housed by the flange. The device also has roller for keeping the space in the rack-gear engagement constant. SUBSTANCE: converter comprises the rack, gearing, output shaft, two overloading clutches, and thrusts. The rack connected with the pivot is in coincidence with the overrunning clutches which are in a staggered contact with conical gears. The movement of the conical pivots are limited by means of the pinion secured to the output shaft. The racks provided with the gearing and overrunning clutches are interposed between the two conical gearings. SUBSTANCE: gear converter comprises nonrotating shaped gear wheel (1) with inner teeth engagement, which reciprocates with respect to semiaxle (3) that rotates together with shaped pinion (2) secured to the semiaxle from one side and cooperating with shaped gear wheel (1). The length of the gear rim of pinion (2) is twice as less as that of gear rim of shaped gear wheel (1). The continuous engagement of the teeth of shaped gear wheel (1) with pinion (2) is provided by means of the cooperation of cam (4) that is connected with pinion (2) with roller (8) freely rotating around its axis (9). SUBSTANCE: gearing device comprises two gear racks and gear wheel. The recess is provided between the racks (26) and (28) to prevent gear wheel (10) against the contact with both of the racks simultaneously. The adapting gear member between two racks is defined by the convex gear arc. When gear wheel (10) engages adapting member (30), it disengages rack (26) immediately before the moment when it engages rack (28). SUBSTANCE: according to invention, body 10 forms prismatic chamber 12 whose cross section is oval of odd order formed from arcs 34, 36, 38 with first smaller radius of curvature and arcs 40, 42, 44 with second, larger radius of curvature changing continuously and differentially one into the other. Thus, corresponding cylindrical parts of inner surface of chamber are formed. Chamber 12 accommodates rotating piston 60 whose cross section forms oval of the order smaller by 1 than order of chamber 12. Opposite parts of side surface are formed on rotating piston 60, one of which rotates in part of inner surface of radius of curvature equal to said part and the other adjoins opposite part of inner surface to slide along surface. rotating piston 60 divides chamber 12 in any position into two working spaces 78, 80. Instantaneous axes of rotation 112, 114 of rotating piston 60 are determined on middle plane of piston being fixed for a short time. Working agent to set rotating piston 60 into motion is periodically introduced into working spaces. Rotating piston 60 rotates in each phase of its motion in one of opposite parts of its side surface 70 in corresponding part of inner side surface 62 of chamber around corresponding instantaneous axis of rotation 112 and slides by opposite part of its surface 72 along corresponding opposite part of inner side surface 54 of chamber 12 to stop, i.e. until it comes into extreme position. Then, to execute following phase of movement, instantaneous axis of rotation jumps from previous position into second possible position 114 relative to piston and is fixed in this position for a short rime. Driven or driving shaft 102 is in engagement with rotating piston 60. To prevent kinematic ambiguity of instantaneous axis of rotation in extreme position, instantaneous axis of rotation is mechanically fixed in each extreme position for a time (Fig.1). FIELD: the invention refers to the field of machine building namely to transformers of alternate/reciprocal motion into rotary and vice-versa and may be used in piston engines, pumps, compressors and other gears. SUBSTANCE: the transformer is a planetary toothed mechanism consisting of a non-rotating central pinion with inner toothed gearing making alternate/reciprocal motion relatively to the axle rotating together with firmly connected with it from one side and interacting with the central pinion-satellite. The length of the toothed rim of the satellite is two times smaller then the length of the toothed rim of the central pinion, of the link or of the mechanism providing continuous engagement of the teeth of the central pinion and the satellite. The link and the mechanism are fulfilled in such a manner that install marginally allowable distance between the positions of their joining with the central pinion and with the satellite not installing the maximum allowable value of this distance. The central pinion and the satellite have figured different profile according to their toothed rims. EFFECT: increases the coefficient of efficiency of transformation of the alternate/reciprocal motion into rotary and vice-versa and simplifies the construction of the toothed transformer. SUBSTANCE: motion converter comprises four cylinders grouped in pairs. Pistons of opposite cylinders are connected with each other through fixed bar. The converter has the first and the second pair of fixedly connected parallel toothed racks brought into engagement with corresponding segmental gear-wheels fixedly connected to one output shaft. In each segmented gear-wheel teeth extend for less than half of 180° pitch circle. Segmented gear-wheels cooperating with corresponding toothed racks are shifted through predetermined angle, preferably 90° angle, one relatively another. SUBSTANCE: invention relates to lever-tooth mechanisms with rack and it can be used to drive mechanisms of slotting machine, automobile windshield wiper, brush cleaner of screens of grain-cleaning machines, in automatic machines, etc. Proposed mechanism contains base, toothed wheel and crank hinge-mounted on base for rotation, and toothed rack engaging with toothed wheel. Lever with guide member is hinge-mounted on axle of toothed wheel. Guide member engages with rear side of toothed rack forming translational pair. Guide member is made in form of roller hinge-mounted for rotation on axle installed on free end of lever provided with thread. Axle of roller is provided with through cross hole to pass threaded end of lever. Roller is held on lever by nuts arranged from outer and inner sides of roller to adjust and fix radial clearance in meshing between rack and toothed wheels. | Mid | [
0.647058823529411,
33,
18
] |
Welcome to FulhamSW6.com The site for Fulham residents - and anyone planning to buy, eat or shop in Fulham. If you are viewing this text, your browser lacks the ability to read frames. Although you will need a browser that can handle frames to use our site, you can idea of the site by following some of the links below. When you're seriously looking to buy or let, you need to know EVERYTHING that's available. That's where we make the difference. We already have over 600 houses for sale or to let in Fulham, listed on our database, which is searchable by price/rent and number of bedrooms. | Low | [
0.45823927765237005,
25.375,
30
] |
[Kinetic model for extraction process of Chinese traditional medicine]. To present a theory bases for Chinese traditional medicine extraction process. A kinetic equation for Chinese traditional medicine extraction process was presented based on Fick's second law of diffusion. Experiments for extracting tanshinone from Salvia miltiorrhiza Bge with ethanol under different particle sizes and temperatures were studied. In addition, the data of extraction process of three kinds of herbs, including Glycyrrhiza uralensis Fisch, Schisandra chinesis and Ophiopogon japonicus (Thunb.) Ker-Gawl were introduced to evaluate the kinetic equation. The kinetic of Chinese traditional medicine extraction process agreed with the first-order rate equation. Experiment data of active components extracted from the herbs above could perfectly match with the kinetic equation. The rate-determining step for extracting tanshinone from Salvia miltiorrhiza Bge was the diffusion of tanshinone through the herb particles. The apparent activation energy for tanshinone diffusion process was 20.26 kJ.mol-1. The relation between the concentration of active component and extraction time, size of herb particles and temperature could be got from this kinetic equation. The extraction process of other herbs besides the above could be described by this equation. | High | [
0.683994528043775,
31.25,
14.4375
] |
Alberta's security regulator is proposing 'disclose or explain' rules as a way to narrow the gender gap in Alberta boardrooms. The new rules will require Alberta-based companies which are listed on major stock exchanges such as the Toronto Stock Exchange, to disclose the number of women on their boards and in executive management. If adopted by the Alberta Securities Commission, publicly traded companies would also be required to disclose any gender-parity policies which inform hiring practices or the selection of executive officers. In 2015, women held only nine per cent of board positions on Alberta-based, TSX-listed companies. The new rules would help shatter the glass ceiling, said Stephanie McLean, Alberta's minister of status of women. "Women today are well educated, trained and have the leadership, business skills and experience to match their male counterparts," she said. 'Too few women at the CEO level' "We have a woman premier, a woman deputy premier, and a woman chief justice, and a woman at the head of Alberta's public service. Yet we still see relatively too few women at the CEO level in corporate Alberta," McLean said. "It's time we achieve gender balance at all decision-making tables in Alberta — including corporate boardrooms — and this is an important step." The amendments would bring Alberta in line with most other jurisdictions in Canada. All provinces and territories, with the exception of Alberta and British Columbia, adopted the disclosure amendments in November 2014. The ASC rejected the rule change in 2014 over questions whether the composition of boards was within the commission's mandate, said chair Stan Magidson Wednesday. "On further reflection, and at the request of the government who asked us to reconsider this, we have taken a fresh look at it," he said. Two years later, the commission sees things differently, Magidson said. "It's not about what the actual composition of your board is or isn't, but it's that you provide sunlight on this matter to shareholders at large from which they can make their investment decisions." The glass ceiling Despite the increased accountability, corporate Canada remains largely impregnable to women. Women make up nearly half of the Canadian workforce, yet only one-quarter hold senior management positions and fewer than 16 per cent serve as directors, the province said. Almost half of companies listed on the Toronto Stock Exchange still have no women on their boards. If adopted, the amendments would require the listed companies, or non-venture issuers, to provide annual disclosure of the following items: policies regarding the representation of women on the board, if any; whether the board or its nominating committee considers the representation of women in the director identification and selection process; whether the company considers the representation of women in executive officer positions when making executive officer appointments; targets regarding the representation of women on the board and in executive officer positions, if any have been set by the issuer; the number of women on the board and in executive officer positions; and director term limits or other mechanisms of board renewal The suggested changes are subject to a 30-day comment period, before a final vote can be made. | High | [
0.656934306569343,
33.75,
17.625
] |
#!/usr/bin/env python # -*- coding: utf-8 -*- import dbus import gobject from base64 import b64encode from os import system from msa import get_logger from msa.core.ipc.server.dbus_service import MSADbusService from msa.core.serial.printer import obtener_impresora from msa.core.settings import DBUS_IMPRESORA_PATH, DBUS_BUSNAME_PRINTER, \ SCAN_DELAY, PATH_IPC_SERVER from msa.helpers import levantar_locales levantar_locales() logger = get_logger("print_service") class PrinterDBus(MSADbusService): def __init__(self): """Constructor""" self.object_path = DBUS_IMPRESORA_PATH self.bus_name = DBUS_BUSNAME_PRINTER self._conn = False MSADbusService.__init__(self, True) def _real_init(self): self.ultimo_estado_tarjeta = None self.printer = obtener_impresora() def _tarjeta_ingresada(callback): try: ingresada = self.printer.tarjeta_ingresada() if not self._conn: self.connection(True) self._conn = True if ingresada is None: ingresada = False if ingresada != self.ultimo_estado_tarjeta: self.ultimo_estado_tarjeta = ingresada callback(self.full_paper_status()) # si falla viendo si hay papel o no asumimos que la impresora esta # desconectada except Exception, e: if self._conn: self._conn = False self.connection(False) return True gobject.timeout_add(SCAN_DELAY, _tarjeta_ingresada, self.insertando_papel) @dbus.service.method(DBUS_BUSNAME_PRINTER) def quit(self): """ Cierra el servicio DBUS, útil para casos de reinicio""" if self._loop.is_running(): self._loop.quit() def _info(self, func): logger.info("llamando a %s" % func) @dbus.service.method(DBUS_BUSNAME_PRINTER) def imprimiendo(self): self._info("imprimiendo") return self.printer.imprimiendo() @dbus.service.method(DBUS_BUSNAME_PRINTER) def get_type(self): self._info("get_type") return self.printer.get_type() @dbus.service.method(DBUS_BUSNAME_PRINTER) def get_vendor(self): self._info("get_vendor") return self.printer.get_vendor() @dbus.service.method(DBUS_BUSNAME_PRINTER) def linefeed(self, n): self._info("linefeed") return self.printer.linefeed(n) @dbus.service.method(DBUS_BUSNAME_PRINTER) def backfeed(self, n): self._info("backfeed") return self.printer.backfeed(n) @dbus.service.method(DBUS_BUSNAME_PRINTER) def expulsar_boleta(self): self._info("expulsar_boleta") self.printer.expulsar_boleta() self.boleta_expulsada() @dbus.service.signal(DBUS_BUSNAME_PRINTER) def boleta_expulsada(self): self._info("boleta_expulsada") return None @dbus.service.method(DBUS_BUSNAME_PRINTER) def tomar_tarjeta(self, loops): self._info("tomar_tarjeta") #dbus no permite pasar None, entonces paso False y convierto if not loops: loops = None return self.printer.tomar_tarjeta(loops) @dbus.service.method(DBUS_BUSNAME_PRINTER) def posicionar_al_inicio(self): self._info("posicionar_al_inicio") return self.printer.posicionar_al_inicio() @dbus.service.method(DBUS_BUSNAME_PRINTER) def limpiar_cola(self): self._info("limpiar_cola") return self.printer.limpiar_cola() @dbus.service.method(DBUS_BUSNAME_PRINTER) def imprimir_image(self, filepath, mode, size, dpi, transpose, compress, only_buffer): self._info("imprimir_image") size = ",".join([str(num) for num in size]) dpi = ",".join([str(num) for num in dpi]) command = "%s/print_or_cache.py" % PATH_IPC_SERVER args = "--filepath %s --mode %s --size %s --dpi %s " % (filepath, mode, size, dpi) if transpose: args += "--transpose " if only_buffer: args += "--only_buffer " #args = list(args.split()) logger.debug("Corriendo comando externo de impresion %s", command) system(command + " " + args) return True @dbus.service.method(DBUS_BUSNAME_PRINTER) def imprimir_serializado(self, tipo_tag, tag, transpose, only_buffer, extra_data): self._info("imprimir_serializado") command = "%s/print_or_cache.py" % PATH_IPC_SERVER args = "--serialized --tipo_tag %s --tag \"%s\" --extra_data \"%s\" " % (tipo_tag, tag, b64encode(extra_data)) if transpose: args += "--transpose " if only_buffer: args += "--only_buffer " logger.debug("Corriendo comando externo de impresion %s", command) logger.info(command + " " + args) system(command + " " + args) return True @dbus.service.method(DBUS_BUSNAME_PRINTER) def do_print(self): self._info("do_print") self.printer.do_print() @dbus.service.method(DBUS_BUSNAME_PRINTER) def simular_impresion(self): self._info("simular_impresion") return self.printer.simular_impresion() @dbus.service.method(DBUS_BUSNAME_PRINTER) def tarjeta_ingresada(self): status = self.full_paper_status() self._info("tarjeta_ingresada") return status @dbus.service.signal(DBUS_BUSNAME_PRINTER) def insertando_papel(self, status): self._info("insertando_papel") return status @dbus.service.method(DBUS_BUSNAME_PRINTER) def full_paper_status(self): self._info("full_paper_status") status = {} try: ingresada = self.printer.tarjeta_ingresada() status['paper_out_1'] = 1 if ingresada else 0 except ValueError, e: status['paper_out_1'] = 0 self._conn = False self.connection(False) return status @dbus.service.method(DBUS_BUSNAME_PRINTER) def tarjeta_sin_retirar(self): self._info("tarjeta_sin_retirar") estado = self.printer.tarjeta_sin_retirar() self._info("estado %s" % estado) status = self.full_paper_status() return status @dbus.service.method(DBUS_BUSNAME_PRINTER) def ping(self): return 'pong' @dbus.service.method(DBUS_BUSNAME_PRINTER) def estado(self, out_signature="b"): return self._conn @dbus.service.signal(DBUS_BUSNAME_PRINTER) def con_tarjeta(self, tiene_tarjeta): self._info("tiene_tarjeta %s" % tiene_tarjeta) status = self.full_paper_status() return status @dbus.service.signal(DBUS_BUSNAME_PRINTER) def connection(self, state): self._info("connection: " + str(state)) return state if __name__ == '__main__': PrinterDBus() | Mid | [
0.567474048442906,
41,
31.25
] |
Sharpen&Bend: recovering curved sharp edges in triangle meshes produced by feature-insensitive sampling. Various acquisition, analysis, visualization, and compression approaches sample surfaces of 3D shapes in a uniform fashion without any attempt to align the samples with sharp edges or to adapt the sampling density to the surface curvature. Consequently, triangle meshes that interpolate these samples usually chamfer sharp features and exhibit a relatively large error in their vicinity. We present two new filters that improve the quality of these resampled models. EdgeSharpener restores the sharp edges by splitting the chamfer edges and forcing the new vertices to lie on intersections of planes extending the smooth surfaces incident upon these chamfers. Bender refines the resulting triangle mesh using an interpolating subdivision scheme that preserves the sharpness of the recovered sharp edges while bending their polyline approximations into smooth curves. A combined Sharpen&Bend postprocessing significantly reduces the error produced by feature-insensitive sampling processes. For example, we have observed that the mean-squared distortion introduced by the SwingWrapper remeshing-based compressor can often be reduced by 80 percent executing EdgeSharpener alone after decompression. For models with curved regions, this error may be further reduced by an additional 60 percent if we follow the EdgeSharpening phase by Bender. | High | [
0.6701846965699201,
31.75,
15.625
] |
Q: string replace using Linq in c# public class Abbreviation { public string ShortName { get; set; } public string LongName { get; set; } } I have a list of Abbreviation objects like this: List abbreviations = new List(); abbreviations.add(new Abbreviation() {ShortName = "exp.", LongName = "expression"}); abbreviations.add(new Abbreviation() {ShortName = "para.", LongName = "paragraph"}); abbreviations.add(new Abbreviation() {ShortName = "ans.", LongName = "answer"}); string test = "this is a test exp. in a para. contains ans. for a question"; string result = test.Replace("exp.", "expression") ... I expect the result to be: "this is a test expression in a paragraph contains answer for a question" Currently I am doing: foreach (Abbreviation abbreviation in abbreviations) { test = test.Replace(abbreviation.ShortName, abbreviation.LongName); } result = test; Wondering if there is a better way using a combination of Linq and Regex. A: If you really wanted to shorten your code, you could use the ForEach extension method on the List: abbreviations.ForEach(x=> test=test.Replace(x.ShortName, x.LongName)); | Mid | [
0.5475113122171941,
30.25,
25
] |
Québec a choisi de prolonger la ligne bleue du métro de Montréal même si les études qu’il a commandées montraient qu’il aurait été plus avantageux de prolonger la ligne orange, a découvert notre Bureau d’enquête. Ces études avaient été gardées secrètes. Elles ont servi de base au conseil des ministres pour aller de l’avant, en avril 2018, avec ce projet évalué à 3,9 milliards $. Bien que l’Autorité régionale de transport métropolitain (ARTM) ait refusé de nous les fournir via une demande d’accès à l’information, nous avons pu les consulter au cours des dernières semaines. Elles ont été réalisées entre 2011 et 2015 par la défunte Agence métropolitaine de transport (AMT), pour analyser divers prolongements des lignes bleue (vers l’est), orange (jusqu’à Laval) et jaune (à Longueuil). Les experts concluent qu’un prolongement de 2,1 km de la ligne orange entre la station Côte-Vertu et la gare de train Bois-Franc aurait, au kilomètre, transporté plus d’usagers, fait abandonner la voiture par plus de gens et permis de meilleurs gains de temps qu’un prolongement de la ligne bleue. Cette option aurait aussi été la seule à soulager le tronçon est de la ligne orange, contrairement au projet de ligne bleue qui risque de le surcharger davantage. «Le petit bout entre Côte-Vertu et Bois-Franc c’est une connexion qui aurait dû être faite depuis bien longtemps. Si j’avais été à l’AMT, j’aurais mis ça comme priorité. En deuxième c’est le métro d’Anjou», estime la professeure associée au Département d’études urbaines de l’Université du Québec à Montréal (UQAM), Florence Junca-Adenot. Selon elle, le prolongement du tronçon ouest de la ligne orange devient encore plus pertinent maintenant que le Réseau express métropolitain (REM) passera à Bois-Franc. «Choix politiques» L’experte rappelle qu’en matière de transports, il y a souvent des «choix politiques». Depuis ces études, la planification des transports en commun dans la région de Montréal est passée dans les mains de l’ARTM. Son porte-parole Simon Charbonneau affirme que la priorisation de la ligne bleue a fait «l’objet d’un consensus des élus des municipalités et du gouvernement du Québec». Les documents que nous avons consultés montrent que les futures stations du prolongement de la ligne bleue seraient peu utilisées par rapport à d’autres stations du réseau. Métro inutile? Selon le professeur en urbanisme à la Faculté de l’aménagement de l’Université de Montréal, Jean-Philippe Meloche, le prolongement de la ligne bleue n’est pas «dans la zone payante» pour utiliser un métro. «Plus je m’approche des zones où le sol vaut cher et où la congestion est très importante, plus les bénéfices du tunnel deviennent importants. Plus je m’éloigne de ces zones, moins le bénéfice est important», indique-t-il. Ainsi, il croit que la densité résidentielle de Saint-Léonard et l’espace de surface peu congestionné de la rue Jean-Talon Est ne justifient pas un métro. Il suggère qu’un train de surface, plus léger et moins cher, aurait pu être une meilleure option. Les experts que nous avons consultés s’entendent toutefois pour dire que le nord-est de l’île est mal desservi et a besoin de transport en commun. «Il y a une grosse quantité de gens qui se déplacent pour aller travailler vers l’est et pour aller travailler vers le centre», rappelle Mme Junca-Adenot. TROP TARD POUR RECULER? Puisque le plan d’affaires n’a pas encore été approuvé par Québec, le projet de la ligne bleue pourrait toujours être annulé. Toutefois, des contrats ont été octroyés pour l’architecture des stations dans les dernières semaines, et Ottawa a confirmé du financement pour ce projet. La ligne orange sera encore plus surchargée Le prolongement de la ligne bleue risque de surcharger davantage la ligne orange. Selon les études que nous avons obtenues, l’ajout de cinq stations sur la ligne bleue augmenterait de 6 % le nombre maximal de passagers à circuler en une heure (charge maximale) sur la ligne orange, en direction Côte-Vertu, en 2031. «On n’est pas capable de prendre ces passagers supplémentaires aujourd’hui et on ne sera pas plus capable en 2031», juge M. Meloche. En 2031, l’AMT prévoit que la charge maximale du cœur de la ligne orange sera de 63 000 passagers le matin, dont 3500 venant du prolongement de la ligne bleue. Les trains Azur pouvant accueillir environ 45 000 passagers par heure par direction, confie un expert à notre Bureau d’enquête. Pour prendre tous les passagers prévus en 2031, il faudrait donc de nouveaux trains ou augmenter la fréquence de passage. Le SRB n’aidera pas Selon la STM, le prolongement de la ligne bleue ne va « pas nécessairement » engorger la ligne orange. «Les usagers sont un peu paresseux. Si tu peux rester dans ton mode, sans transférer, tu vas le faire. Le SRB Pie-IX, c’est à l’étude pour qu’il continue jusqu’au centre-ville», donne en exemple M. Chamberland. Mais selon les simulations menées par l’AMT, le Service rapide par bus (SRB) entraînera plutôt «une hausse sur les points de charge maximale de la ligne orange en direction Côte-Vertu». Et sa prolongation jusqu’au centre-ville aurait un «impact limité» sur l’achalandage de la ligne orange, peut-on lire. Jean-Philippe Meloche estime que le REM, qui sera connecté à la ligne bleue, pourrait venir régler ce problème. «Les gens de Saint-Léonard vont pouvoir se rendre au centre-ville, en évitant la ligne orange. Le prolongement de la ligne bleue sans le REM, ça aurait été encore plus mauvais», soutient-il. ACHALANDAGE PROJETÉ EN 2031 Prolongement de la ligne bleue (option retenue) Pie-IX: 6900 Anjou: 5600 Lacordaire: 3000 Langelier: 2900 Viau: 1700 Prolongement de la ligne orange (option écartée) Bois-Franc: 14 100 Poirier: 3300 Les stations avec le plus d’entrées en 2019 Longueuil: 14 500 Honoré-Beaugrand: 9700 Bonaventure: 8900 En nombre d’entrées en période de pointe du matin (PPAM) 4 AUTRES SCÉNARIOS ÉTUDIÉS Une boucle avec la ligne verte Ce scénario consistait à rattacher la ligne bleue à la ligne verte par la station Radisson. Ce projet, plus long et plus complexe, aurait amené beaucoup plus de gens aux nouvelles stations, surtout des usagers actuels du transport en commun. Finir au nord de l’A40 Construire le terminus du métro le long de l’A25, entre Jarry et Bombardier aurait permis d’ajouter un large stationnement pour attirer davantage d’automobilistes sur la ligne bleue. L’achalandage aurait augmenté un peu, mais pas de manière significative. Passer un peu plus au sud, sous la rue Bélanger Cette option de même longueur n’amenait pratiquement aucun changement sur l’achalandage, mais était meilleure pour l’accès en transport actif. Un prolongement moins long jusqu’à Lacordaire Étonnamment, le tracé plus court scruté en 2013 allait jusqu’à Lacordaire et non jusqu’à Pie-IX comme le projet annoncé par le gouvernement péquiste dans les années 1990. Cette option aurait causé un problème de rabattement des autobus à cause du peu d’espace non bâti au coin des rues Lacordaire et Jean-Talon. | Mid | [
0.6187050359712231,
32.25,
19.875
] |
Q: A Problem With FileZilla This is urgent. I am using FileZilla in order to connect to a server. However, no matter how much I try, it just keeps telling me, "Critical error. Could not connect to server." I am really frustrated at the fact that FileZilla would not let me do something as simple as this. What might be the most probable reason as to why it wouldn't let me connect? For the "host" part, I entered the domain name, without "www". For the port, I entered nothing, assuming that the default port will be used. However, there is nothing I can do unless I am connected to the server. I would appreciate your advice very much. Thank you very much in advance. Filezilla Log: Status: Resolving address of www.pdaphouston.org Status: Connecting to 72.167.232.31:21... Status: Connection established, waiting for welcome message Response: 220---------- Welcome to Pure-FTPd [privsep] [TLS] ---------- Response: 220-You are user number 2 of 500 allowed. Response: 220-Local time is now 13:07. Server port: 21. Response: 220-This is a private system - No anonymous login Response: 220 You will be disconnected after 3 minutes of inactivity. Command: USER <username they provided> Response: 331 User <username they provided> OK. Password required Command: PASS ******** Response: 530 Login authentication failed Error: Critical error Error: Could not connect to server A: From the comments in the question it shows that you have the wrong username and password This is the message: 530 Login authentication failed Check the login details and try again. | High | [
0.6589327146171691,
35.5,
18.375
] |
Q: Locking sub-directories from the view of certain users I have certain sub-dirs that I do not want certain developers to have access to when they login via FTP. They need to view most files/dirs, but some are off limits. Is there any way to block viewing these sub-directories by certain users when they login via FTP but not prevent PHP and Apache from including required files? This is a Linux server that runs cPanel and WHM. A: Yes. You can do this. Use standard POSIX permissions. | Mid | [
0.6334231805929921,
29.375,
17
] |
Exploring the role of motivational interviewing in adolescent patient-provider communication about type 1 diabetes. Adolescents with type 1 diabetes (T1D) frequently experience deterioration in glycemic control. Providers have unique opportunities to address diabetes self-management, yet little is known about the most effective way to communicate with adolescents. This investigation used a motivational interviewing (MI) framework to characterize naturally-occurring adolescent patient-provider communication in medical encounters and examined relations between adolescent patient-provider communication and (a) T1D self-management and (b) glycemic control (hemoglobin A1c [HbA1c]). Medical encounters between pediatric endocrine providers and 55 adolescents with T1D (49% female; M age = 14.8 years; M baseline HbA1c = 8.6%) were audio recorded and coded using standardized rating instruments. Patients and parents completed measures assessing T1D care behaviors and self-efficacy. Assessments were completed at routine endocrinology visits (baseline) and 1 and 3-month post-baseline; HbA1c was obtained from medical records at baseline and 3-month. Hierarchical multiple regressions showed that greater provider use of MI non-adherent behaviors (eg, confronting, persuading) was associated with (a) poorer 3-month HbA1c, P < 0.001; (b) worse 1-month adolescent diabetes adherence P < 0.001, and (c) lower diabetes self-efficacy at 1-month (P < 0.001) follow-up. Lower patient self-efficacy for diabetes self-management mediated the relation between provider use of MI non-adherent language and lower diabetes adherence (P = 0.020). Providers' use of persuasion and confrontation regarding risks of non-adherence was associated with poorer glycemic control and adherence. Communication training for providers that targets reductions in MI-inconsistent language may have the potential to improve diabetes self-care in this vulnerable population. | High | [
0.7291666666666661,
30.625,
11.375
] |
--- abstract: 'We prove that quasi-morphisms and quasi-states on a closed integral symplectic manifold descend under symplectic reduction to symplectic hyperplane sections. Along the way we show that quasi-morphisms that arise from spectral invariants are the Calabi homomorphism when restricted to Hamiltonians supported on stably displaceable sets.' address: 'Department of Mathematics, University of Chicago, Chicago, Illinois 60637' author: - Matthew Strom Borman bibliography: - 'symplectic.bib' title: | Symplectic reduction of\ quasi-morphisms and quasi-states --- [^1] Introduction and results ======================== An overview ----------- In the series of papers [@EntPol03; @EntPol06; @EntPol08; @EntPol09RS], Entov and Polterovich introduced a way to construct quasi-morphisms on the universal cover of the Hamiltonian group and symplectic quasi-states, when $(M, \w)$ is a closed symplectic manifold. In order to determine if their construction will work for a specific $(M^{2n}, \w)$, one must compute the quantum homology ring $QH_{2n}(M,\w)$ and determine if it contains a field summand, which in practice can be a nontrivial task. In this paper we introduce a reduction method that allows one to build quasi-morphisms and symplectic quasi-states from known examples, without requiring further quantum homology calculations. This process is presented formally in Theorems \[thm: descent qm\] and \[thm: descent\], which are the main results of this paper. As an example, suppose that $(M, \w; \Si)$ forms a subcritical polarization where $\Si \subset M$ is a complex hypersurface. If $\z\co C^\infty(M) \to \R$ is a symplectic quasi-state that vanishes on functions with displaceable support, then our construction produces a symplectic quasi-state $\bar{\z}\co C^\infty(\Si) \to \R$. The main tool of the construction is Biran’s decomposition theorem [@Bir01], which roughly states that the unit disk normal bundle $E_\Si$, for the hypersurface $\Si \subset M$, is symplectomorphic to $M \setminus \Del$, where in the subcritical case $\Del$ is a displaceable set (see Section \[Biran\] for a precise statement). This allows us to define a map $\T\co C^\infty(\Si) \to C^\infty(M)$, which lifts functions $H$ on $\Si$ to $E_\Si$, multiplies them by a function that vanished at the boundary of the fibers, and then views the result as a function on $M$ using Biran’s symplectomorphism. The reduced symplectic quasi-state $\bar{\z}\co C^\infty(\Si) \to \R$ is the result of pulling $\z$ back by the map $\T$. The construction for quasi-morphisms proceeds in a similar fashion. This reduction method produces quasi-morphisms and symplectic quasi-states on the Hirzebruch surfaces, which were shown to have quasi-morphisms and quasi-states by Ostrover [@Ost06], but required quantum homology calculations. These and other examples are discussed in Sections \[subcritical\] and \[critical\]. The Hamiltonian group --------------------- Let $(M^{2n}, \w)$ be a closed symplectic manifold, any smooth function $H \in C^\infty(M)$ determines a Hamiltonian vector field $\operatorname{sgrad}H$ on $M$ by $\w(\operatorname{sgrad}H, \cdot) = -dH$. In this manner any time-dependent Hamiltonian $F\co M \times [0,1] \to \R$ gives a Hamiltonian isotopy $\phi_F = \{f_t\}_{t \in [0,1]}$, by solving the differential equation $\d_t f_t = (\operatorname{sgrad}F_t)_{f_t}$ with initial condition $f_0 = {\boldsymbol{\mathbbm{1}}}$. The collection of all such time one maps $f_1$ is the *Hamiltonian group* $Ham(M,\w)$. Seen as a Lie group, its Lie algebra is $\H(M)$, all $H \in C^\infty(M)$ that are *normalized* to have mean zero $\int_M H \w^n = 0$, and the bracket is the Poisson bracket $$\{H, K\} = \w(\operatorname{sgrad}K, \operatorname{sgrad}H) = dH(\operatorname{sgrad}K).$$ Any smooth path in $Ham(M)$ is a Hamiltonian isotopy, so the space of smooth paths based at the identity $\PH(M)$, can be identified with $\P\H(M)$, the space of functions $F\co M \times [0,1] \to \R$ such that $F_t \in \H(M)$ at all times. The group structure of time-wise product on $\PH(M)$ carries over to $\P\H(M)$ as $\phi_F\phi_G = \phi_{F\#G}$ and $\phi_F^{-1} = \phi_{\bar{F}}$ where $$(F\#G)(x,t) = F(x,t) + G(f_t^{-1}(x),t) \quad \mbox{and} \quad \bar{F}(x,t) = -F(f_t(x),t).$$ The universal cover $\CH(M)$ is the path space $\PH(M)$ where elements are considered up to homotopy with fixed endpoints. See [@McDSal98; @Pol01] for further preliminaries and note we use the sign conventions of [@Pol01]. Quasi-morphisms on $\CH(M)$ --------------------------- A *homogeneous quasi-morphism* on a group $G$ is a function $\mu\co G \to \R$ so that $n\mu(g) = \mu(g^n)$ for all $n \in \Z$ and $g \in G$, and for some $C \geq 0$: $$\label{quasi} {\left| \mu(g_1 g_2) - \mu(g_1) - \mu(g_2) \right|} \leq C \quad \mbox{for all $g_1, g_2 \in G$}.$$ The smallest such $C$ is the *defect* and is denoted $D(\mu)$. Homogeneous quasi-morphisms are conjugation invariant and they are genuine homomorphisms when restricted to abelian subgroups. On perfect groups, like $Ham(M)$ and $\CH(M)$, they are the best one can hope for in terms of a map to $\R$. If $\mu$ only statisfies , then $\mu$ is a *quasi-morphism* and can be homogenized by forming $\check{\mu}(g) = \displaystyle{\lim_{n \to \infty}\tfrac{\mu(g^n)}{n}}$, which will be the unique homogeneous quasi-morphism within a bounded distance from $\mu$. See [@Cal09; @Kot04] for more information about quasi-morphisms. In [@EntPol03], Entov and Polterovich introduced a way to build homogeneous quasi-morphisms $\mu\co\CH(M) \to \R$ using spectral invariants from Hamiltonian Floer theory. We will briefly recall the construction and refer the reader to [@EntPol03; @EntPol08; @EntPol09RS] for more details. Let $QH_{*}(M,\w)$ be the quantum homology ring for $(M, \w)$, which is $H_*(M; \C) \otimes_\C \La_\w$ where $\La_\w$ is a Novikov ring associated to $(M,\w)$, and its ring structure is given by a deformation of the normal intersection product using Gromov-Witten invariants. To each element $a \in QH_*(M)$ there is an associated spectral invariant, defined in terms of Hamiltonian Floer theory, which is a functional $c(a, \cdot)\co C^\infty(M \times [0,1]) \to \R$. With $c(a, \cdot)$, one can define the functional $\mu(a, \cdot)\co C^\infty(M \times [0,1]) \to \R$: $$\label{qm def} \mu(a, F) = \int_0^1\int_M F(x,t)\, \w^n dt -\operatorname{Vol}(M) \lim_{k\to \infty} \frac{c(a,F^{\#k})}{k},$$ which descends to a function $$\mu(a, \cdot)\co \CH(M) \to \R.$$ If $a \in QH_{2n}(M)$ is an idempotent and $a * QH_{2n}(M)$ is a field [@EntPol09RS Remark 1.23], then $\mu(a, \cdot)$ is a homogeneous quasi-morphism. Recently Usher [@Ush10 Theorem 1.4] and Fukaya, Oh, Ohta, and Ono [@FukOhOht11a Theorem 1.2] have independently extended this construction using idempotents in the big quantum homology ring, and the associated spectral invariants. Any quasi-morphism build using spectral invariants we will call a *spectral quasi-morphism*. Spectral quasi-morphisms have the following two additional properties, where $\CH_U(M)$ are all Hamiltonian paths generated by Hamiltonians with support in $U$ (not necessarily normalized): 1. *Calabi Property:* If $U \subset M$ is open and displaceable, then $\mu$ restricted to $\CH_U(M)$ is the Calabi homomorphism $\operatorname{Cal}_U$ given by $$\operatorname{Cal}_U(\phi_F) = \int_0^1\int_U F_t\, \w^n dt.$$ 2. *Stability Property:* There is some $B > 0$, so that for all $F, G \in \P\H(M)$, $$\int_0^1 \min_M(F_t - G_t)\,dt \leq \frac{\mu(\phi_G) - \mu(\phi_F)}{B} \leq \int_0^1 \max_M(F_t-G_t)\,dt.$$ In the case of spectral quasi-morphisms $B$ can be taken to be $\operatorname{Vol}(M)$. We will say that a quasi-morphism $\mu$ on $\CH(M)$ is *stable* if it has the stability property. A subset $X \subset M$ is *displaceable* if there is a $\vp \in Ham(M,\w)$ so that $\vp(X) \cap \overline{X} = \emptyset$ and $X$ is *stably displaceable* if $X \times S^1 \subset M \times T^*S^1$ is displaceable. Stable displaceability does not imply displaceability as shown in [@EntPol09RS Example 1.30]. For spectral quasi-morphisms, the Calabi property can be strengthen to include stably displaceable sets, which will be proved in Section \[calabi\]. \[thm: strong calabi\] Let $U \subset (M, \w)$ be a stably displaceable open subset, then spectral quasi-morphisms $\mu(a, \cdot)\co \CH(M) \to \R$ restrict to the Calabi homomorphism $\operatorname{Cal}_U$ on $\CH_U(M)$. Recently Shelukhin [@She11 Corollary 1] proved that there is a quasi-morphism on $\CH(M)$ for any closed symplectic manifold, using a construction that is not Floer-theoretic. Theorem \[thm: descent qm\] and \[thm: descent\] below cannot be directly applied to Shelukhin’s quasi-morphisms since they do not have the Calabi property and they do not induce quasi-states. Symplectic quasi-states and quasi-measures ------------------------------------------ *Symplectic quasi-stats* are functionals $\z\co C^\infty(M) \to \R$ such that for $H,K \in C^\infty(M)$, $a \in \R$: 1. *Quasi-linearity*: If $\{H, K\} = 0$, then $\z(H + aK) = \z(H) + a\,\z(K)$. 2. *Monotonicity*: If $H \leq K$, then $\z(H) \leq \z(K)$. 3. *Normalization*: $\z(1) = 1$. Note that these properties imply that $\z$ is Lipschitz with respect to the $C^0$-norm, ${\left| \z(H) - \z(K) \right|} \leq {\left\| H-K \right\|}$. This notion was introduced by Entov and Polterovich [@EntPol06] and is the symplectic version of Aarnes’ notion of a *topological quasi-state* [@Aar91]. These are functionals $\z\co C(X) \to \R$, for $X$ a compact space, that are monotonic, normalized, and quasi-linear in the sense that if $F \in C(X)$ and $g,h\in C(\R)$, then $\z(g\circ F + h\circ F) = \z(g\circ F) + \z(h\circ F)$. Aarnes proved a Riesz representation theorem for quasi-states showing they correspond to topological quasi-measures, which are additive set functions $\t\co \mathcal{A} \to [0,1]$ where $\mathcal{A}$ are the subsets of $X$ that are either open or closed. The quasi-state $\z$ determines the quasi-measure $\t_\z$, where for a closed set $C \subset X$ $$\t_\z(C) = \inf\{\z(F) \mid F: X \to [0,1],\, F \geq {\boldsymbol{\mathbbm{1}}}_C\}.$$ Being a symplectic quasi-state is a stronger condition than being a topological quasi-state. A stable homogeneous quasi-morphism $\mu \co \CH(M) \to \R$ induces a symplectic quasi-state $\z_\mu \co C^\infty(M) \to \R$ via $$\label{qs from qm} \z_\mu(H) = \frac{\int_M H \w^n - \mu(\phi_{H_n})}{\operatorname{Vol}(M)},$$ where $H_n = H - \tfrac{\int_M H \w^n}{\operatorname{Vol}(M)}$. Quasi-linearity of $\z_{\mu}$ follows from the fact that $\p_H$ and $\p_K$ commute if $\{H, K\} = 0$, and quasi-morphisms are homomorphisms on abelian subgroups. The monotonicity of $\z_{\mu}$ follows from the stability property for $\mu$. Symplectic quasi-states formed as in are $Ham(M)$ invariant, since $\mu(\phi_H) = \mu(\psi^{-1} \phi_H \psi)$ by conjugation invariance of quasi-morphisms, and $\psi^{-1} \phi_H \psi = \phi_{H\circ \psi}$. If $\mu$ has the Calabi property, then $\z_\mu$ vanishes on functions with displaceable support and hence the associated quasi-measure $\t_\mu$ will vanish on displaceable sets. Since the symplectic quasi-state $\z_{\mu}$ is $Ham(M)$ invariant, the associated quasi-measure $\t_\mu$ will also be $Ham(M)$ invariant, which allows it to detect nondisplaceability. For if $\vp\in Ham(M)$ displaced $X \subset M$, then $\tau_\mu(X) \leq 1/2$ since $$2\, \tau_\mu(X) = \tau_\mu(X) + \tau_\mu(\vp(X)) = \tau_\mu(X \cup \vp(X)) \leq 1.$$ Consider now a spectral quasi-morphism $\mu(a, \cdot): \CH(M) \to \R$. The associated symplectic quasi-state $\z(a, \cdot): C^\infty(M) \to \R$ is given by $$\z(a, H) = \lim_{k \to \infty} \frac{c(a, kH)}{k},$$ and we can form the associated symplectic quasi-measure $\t(a, \cdot)$. In [@EntPol09RS Definition 1.4], the notion of a closed subset $X \subset M$ being *(super)heavy* (with respect to an idempotent $a$) was introduced. When $a$ gives a spectral quasi-morphism, these two notions agree and are equivalent to $\t(a, X) = 1$ [@EntPol09RS Remark 1.23]. In this language they prove [@EntPol09RS Theorem 1.4] that if $X$ is stably displaceable, then $\t(a,X) \not= 1$. This result can be strengthen to the following theorem, which is proven in Section \[proofs2\]. \[thm: stably\] For any subset $V \subset M$, if $V$ is stably displaceable, then $\tau(a, V) = 0$. For a symplectic quasi-state $\z\co C^\infty(M) \to \R$, consider the inequality $$\label{master} \Pi_\z(H,K) := {\left| \z(H + K) - \z(H) - \z(K) \right|} \leq C \sqrt{{\left\| \{H,K\} \right\|}}.$$ In [@EntPolZap07 Theorem 1.4] it is proved that holds for symplectic quasi-states formed from stable homogeneous quasi-morphisms as in , where $C$ can be taken to be $\sqrt{2D(\mu)}/\operatorname{Vol}(M)$. If holds for some $C$, let $C(\z)$ be the smallest number for which it is satisfied for all $H, K \in C^\infty(M)$. Due a result of Cardin and Viterbo [@CarVit08 Theorem 1.2], which was later generalized by Entov and Polterovich [@EntPol10 Theorem 1.1] and Buhovsky [@Buh10 Theorem 1.1.2], it makes sense to say that two continuous functions $H$ and $K$ Poisson commute if they have smooth $C^0$-approximations $H_n$ and $K_n$ such that ${\left\| \{H_n, K_n\} \right\|} \to 0$. Therefore says that symplectic quasi-states are linear on continuous functions that Poisson commute. The inequality is the main tool used to lower bound the Poisson bracket invariants introduced by Buhovsky, Entov, and Polterovich [@BuhEntPol11]. Polarizations and Biran’s decomposition theorem {#Biran} ----------------------------------------------- A *polarization* $\P=(M^{2n},\O, J;\Si)$ is a closed Kähler manifold with $[\O] \in H^2(M,\Z)$ and a closed connected complex hypersurface $\Si \subset M$ such that $[\Si] \in H_{2n-2}(M, \Z)$ is Poincaré dual to $k[\O] \in H^2(M, \Z)$. By scaling $\O$ by $k$, we will assume that $k = 1$ and hence $\operatorname{Vol}(M, \O) = \operatorname{Vol}(\Si, \O_\Si)$. Given a polarization $\P= (M, \O, J; \Si)$, there is a compact isotropic celluar subspace $\Del_{\P} \subset M$, which is disjoint from $\Si$ and is called the *skeleton* associated to $\P$. An example of a polarization is a linear $\CP^{n-1} \subset \CP^n$, where $\CP^n$ is given the Fubini-Study form $\O$ with $\int_{\CP^1} \O = 1$. If $\CP^{n-1}$ is given by $\{z_0 = 0\}$, then the skeleton $\Del = \{[1\pc 0 \pc \cdots \pc 0]\}$ is a single point. One can build a symplectic disk bundle $\pi\co (E_\Si, \w_{\operatorname{can}}) \to \Si$ using the normal bundle from $\Si \subset M$, such that $(\Si, \w_\Si)$ symplectically embeds into $(E_\Si, \w_{\operatorname{can}})$ as the zero section. Biran’s decomposition theorem for polarizations [@Bir01 Theorem 2.6.A] says there is a canonical symplectomorphism $F_\P$ such that the following diagram commutes: $$\label{decom} \xymatrix{ (E_\Si, \w_{\operatorname{can}}) \ar[r]^{F_\P} & (M \setminus \Del_\P, \O)\\ (\Si, \O_\Si) \ar[u]^{\mbox{$0$-section}} \ar@{=}[r] & (\Si, \O_\Si) \ar[u]_{\mbox{inclusion}} }$$ The disk bundle $E_\Si$ comes with a hermitian metric on the fibers such that each fiber is the open disk of radius one. The radius on the fibers gives a global coordinate function $r$ on $E_\Si$ and the map $F_\P$ is such that any open neighborhood of $\Del_\P$ contains the complement of $F_\P(\{e \in E_\Si \mid r(e) \leq 1-\e\})$ for some small $\e$. For more details see Section \[setup\] and [@Bir01; @Bir06; @BirCie01]. Symplectic reduction of quasi-morphisms and symplectic quasi-states ------------------------------------------------------------------- We will now describe our procedure for reducing quasi-morphisms and symplectic quasi-states from the total space $M$ of a polarization to the hypersurface $\Si$, which is the main idea in this paper. Let $\th\co [0,1] \to \R$ be a nonnegative function such that $\th(0) = 1$, vanishes in a neighborhood of $1$, and defines a smooth function $\th(r)$ on $E_\Si$, where $r$ is the radial coordinate on $E_\Si$. To $\th$ we associate a linear, order preserving map $$\label{Theta} \T\co C^\infty(\Si) \to C^\infty(M) \quad \mbox{given by} \quad \T(H) = (F_\P^{-1})^*(\th \cdot \pi^*H),$$ where $\pi\co E_\Si \to \Si$ is the projection and $F_\P$ is the symplectomorphism in . When notationally convenient we will consider $\th$ as the the function $\T(1)$ on $M$, and observe that the condition that $\th$ vanishes near $1$ ensures that $\T(H)\co M \to \R$ is smooth. If $H \in C^\infty(\Si)$ has zero mean, then so does $\th\cdot\pi^*H \in C^\infty(E_\Si)$ (see ). Therefore we can view $\T$ as a linear map $$\T\co \P\H(\Si) \to \P\H(M).$$ A nice class of maps are given by $\T_\e$, where $\th_\e(r)\co [0,1] \to \R$ interpolates between $1-r^2$ when $r \leq 1-\e$ and zero when $r \geq 1-\tfrac{\e}{2}$. \[thm: descent qm\] Let $\mu\co \CH(M) \to \R$ be a homogeneous quasi-morphism and suppose there is an open neighborhood $U$ of $\Del_\P$ so that $\mu$ restricts to the Calabi homomorphism on $\CH(M)_U$. Then $\mu$ descends to a homogeneous quasi-morphism $\hat{\mu}$, where for any $\e$ sufficiently small: $$\label{qmorph descent} \hat\mu\co \CH(\Si) \to \R \quad \mbox{is defined by} \quad \hat\mu(\vp) = \mu(\phi_{\T_\e(F)}),$$ where $F \in \P\H(\Si)$ generates $\vp \in \CH(\Si)$. The defect of $\hat\mu$ has upper bound $D(\hat\mu) \leq 2D(\mu)$. If $\mu$ is stable, then so is $\hat\mu$, and if $\mu$ also has the Calabi property then the normalized version of $\hat\mu$ $$\bar\mu\co \CH(\Si) \to \R \quad\mbox{defined by} \quad \bar\mu(\vp) = \z(\th_\e)^{-1} \mu(\phi_{\T_\e(F)})$$ will also have the Calabi property, where $\z$ is the symplectic quasi-state associated to $\mu$. The condition on $\e$ is that the complement of $F_\P(\{e \in E_\Si \mid r(e) \leq 1-\e\})$ is a subset of the neighborhood $U$ of $\Del_\P$. Since $\mu$ is the Calabi homomorphism on $\CH(M)_U$, it follows from that $\z$ vanishes on functions supported in $U$. Therefore $M \setminus U$ has full quasi-measure $\tau_{\z}(M \setminus U) = 1$, and hence $\z(\th_{\e}) > 0$ since $\th_{\e}$ is strictly positive on $M \setminus U$. The map $\T$ can also be used to pullback symplectic quasi-states $\z$ on $M$ to $\Si$. \[thm: descent\] Suppose that $\z\co C^\infty(M) \to \R$ is a symplectic quasi-state and let $\th$ as in be such that $\z(\th) > 0$. Then the functional $\bar\z_\th\co C^\infty(\Si) \to \R$, defined as $$\label{qs reduction} \bar\z_\th(F) = \frac{\z(\T(F))}{\z(\th)},$$ is a symplectic quasi-state. The properties of vanishing on functions with displaceable support and being $Ham$ invariant descend to $\bar\z_\th$ should $\z$ have them. If $\z$ satisfies , then $\bar\z_\th$ does as well with $C(\bar\z_\th) \leq \frac{C(\z)}{\z(\th)} {\left\| \tfrac{\th(r)}{\sqrt{1-r^2}} \right\|}$. Such $\th$ exist if and only if the skeleton $\Del_{\P}$ does not have full quasi-measure with respect to $\t_{\z}$, and in these cases they will exist with abundance. Theorems \[thm: descent qm\] and \[thm: descent\] are proved in Section \[proofs\], and examples of when they apply are given in Sections \[subcritical\] and \[critical\]. If Theorem \[thm: descent qm\] applies to a stable $\mu: \CH(M) \to \R$, then we have two ways to build symplectic quasi-states on $\Si$. We can reduce $\mu$ to $\bar\mu$ as in Theorem \[thm: descent qm\], and then form the associated quasi-state $\z_{\bar\mu}$. Alternatively we can first form the associated quasi-state $\z = \z_\mu$ on $M$, and then reduce it to $\bar\z_{\th_\e}$ as in Theorem \[thm: descent\], where $\e$ must be small enough for Theorem \[thm: descent qm\] to apply. These two symplectic quasi-states $\z_{\bar\mu}$ and $\bar\z_{\th_\e}$ are equal, which can be verified by checking on normalized Hamiltonians $F \in C^{\infty}(\Si)$ $$\z_{\bar\mu}(F) = -\frac{\bar\mu(\phi_{F})}{\operatorname{Vol}(\Si)} = -\frac{\mu(\phi_{\T_{\e}(F)})}{\z(\th_{\e}) \operatorname{Vol}(\Si)} = \frac{\z(\T_{\e}(F))}{\z(\th_{\e})} = \bar\z_{\th_{\e}}(F).$$ Theorems \[thm: descent qm\] and \[thm: descent\] only rely on the Biran decomposition , for which the condition that $(M, \O, J)$ be Kähler is not essential. Let $(M^{2n}, \O)$ be a closed symplectic manifold with $[\O]\in H^{2}(M; \Z)$ and let $\Si^{2n-2}$ be a closed symplectic submanifold Poincaré dual to $k[\O]$ for $k \in \N$. In this setting $(M, \O; \Si)$, Opshtein proved a Biran decomposition result [@Ops11 Theorem 1] where the the skeleton $\Delta$ in has zero volume. Under the additional assumption that $M \setminus \Si$ is a Weinstein manifold, $\Si$ is called a *symplectic hyperplane section* and for these Biran and Khanevsky [@BirKha11 Section 2.3] explain that Biran’s decomposition holds with the skeleton $\Delta$ being an isotropic cellular subspace. Closed symplectic submanifolds $\Si^{2n-2} \subset (M^{2n}, \O)$ that are Poincaré dual to $k[\O] \in H^{2}(M; \Z)$ exist for sufficiently large $k \in \N$ by Donaldson [@Don96]. For $k$ sufficiently large, these Donaldson symplectic hypersurfaces also satisfy the condition that $M \setminus \Si$ is Weinstein by Giroux [@Gir02 Proposition 11]. Suppose that in Theorem \[thm: descent qm\] we replaced Biran’s symplectomorphism $F_{\P}$ in with a symplectic embedding of a disk bundle over $\Si$ into a tubular neighborhood of $\Si$ in $M$. For a small enough tubular neighborhood of $\Si$, reducing any spectral quasi-morphism $\mu(a, \cdot)$ this way will result in the trivial quasi-morphism by Corollary \[cor: coiso\]. Therefore a priori Theorems \[thm: descent qm\] and \[thm: descent\] cannot be made into local constructions. In [@Bor11a], the constructions in this paper are adjusted to the case of symplectic reduction for Hamiltonian $\mathbb{T}^{k}$-actions on a level set of the moment map $\Phi: (M, \w) \to \R^{k}$ with full quasi-measure. In contrast to this paper, the constructions in [@Bor11a] are truly local in nature. This method of reducing quasi-morphisms and symplectic quasi-states also works in the setting of Liouville domains $(M, \w, L)$, when the Reeb vector field for $\a = (\i_L\w)|_{\d M}$ induces a free $S^1$-action on $\d M$. Then the analogues of Theorems \[thm: descent qm\] and \[thm: descent\] hold and allow one to reduce quasi-morphism and symplectic quasi-states on $M$ to the reduction $\Si = \d M/S^1$. Acknowledgments {#acknowledgments .unnumbered} --------------- I would like to thank my advisor Leonid Polterovich for his wonderful help, guidance, and encouragement related to this work. I am grateful to Paul Biran for his generous help with examples of subcritical polarizations and explaining his work with Yochay Jerby to me. I would also like to thank Egor Shelukhin for useful discussions. Remarks and examples ==================== A criterion for a spectral quasi-morphism to restrict to the Calabi homomophism {#calabi} ------------------------------------------------------------------------------- From the definitions, whenever a stable homogeneous quasi-morphism on $\CH(M)$ restricts to the Calabi homomorphism on $\CH_U(M)$ for an open set $U$, the associated quasi-measure vanishes on $U$. It turns out that the converse is true as well for spectral quasi-morphisms. \[prop: zero measure\] Let $\mu(a, \cdot)\co \CH(M) \to \R$ be a spectral quasi-morphism and let $\t(a,\cdot)$ be its associated symplectic quasi-measure. If an open subset $U \subset M$ has zero measure $\tau(a,U) =0$, then $\mu$ restricts to the Calabi homomorphism on $\CH(M)_U$. Let $\vp \in \CH_U(M)$ be generated by a Hamiltonian $F\co M \times [0,1] \to \R$ such that $\operatorname{supp}(F_t) \subset U$ for all $t$. The fact that $\tau(a,U) = 0$ implies that $\z(a,H) = 0$ for all $H \in C^\infty(M)$ that are zero outside of $U$. There exist functions $A, B\co M \to \R$ that are zero outside $U$ so that as time dependent functions $kA \leq F^{\#k} \leq kB$ and hence by the monotonicity of spectral invariants $$0 = \z(A) = \lim_{k \to \infty} \frac{c(a, kA)}{k} \leq \lim_{k \to \infty} \frac{c(a, F^{\#k})}{k} \leq \lim_{k \to \infty} \frac{c(a, kB)}{k} = \z(B) = 0.$$ Therefore the limit term vanishes in and hence $\mu(a, \vp) = \operatorname{Cal}_U(\vp)$. We can now prove Theorem \[thm: strong calabi\]. By Theorem \[thm: stably\], a stably displaceable set $U$ has zero measure and hence by Proposition \[prop: zero measure\], $\mu(a, \cdot)$ restricts to the Calabi homomorphism on $\CH_U(M)$. A corollary to Theorem \[thm: strong calabi\] is the following. \[cor: coiso\] Let $\Sigma \subset (M, \w)$ be a closed nowhere coisotropic submanifold, then $\Sigma$ is stably displaceable and hence its has a neighborhood on which $\mu(a, \cdot)$ restricts to the Calabi homomorphism. A special case being when $\Sigma \subset M$ is a closed symplectic submanifold. Observe that $\Sigma \times S^1 \subset M \times T^*S^1$ is nowhere coisotropic and has at least one nonvanishing section of its normal bundle. Therefore by Gürel’s Displacement Principle [@Gur08 Theorem 1.1], we know that $\Sigma \times S^1$ is displaceable in $M \times T^*S^1$. Reduction for subcritical polarizations. {#subcritical} ---------------------------------------- The skeleton $\Delta_\P$ of a polarization $\P = (M^{2n}, \O, J; \Si)$ is an isotropic cellular subspace and can be replaced with an isotropic CW-complex $\Del$ of the same dimension such that the decomposition still holds [@Bir01 Theorem 2.6.C]. A polarization is *subcritical* if $\dim \Del_\P < n$ and is *critical* if $\dim \Del_\P = n$, which are the only options since $\Del_\P$ is isotropic. In the subcritical case, the Isotopy Theorem [@BirCie01 Theorem 6.1.1] states that $\Del$ is displaceable. The following is now a corollary of Theorems \[thm: descent qm\] and \[thm: descent\]. \[cor: subcritical\] Let $\P = (M^{2n},\O, J;\Si)$ be a subcritical polarization. Calabi quasi-morphisms $\mu\co \CH(M) \to \R$ descend to $\bar\mu\co \CH(\Si) \to \R$ as in Theorem \[thm: descent qm\]. Symplectic quasi-states $\z$ on $(M, \w)$ with the vanishing property descend to symplectic quasi-states $\bar{\z}$ on $(\Si, \O_{\Si})$ as in Theorem \[thm: descent\]. ### The Hirzebruch surfaces are hypersurfaces in subcritical polarizations. The Hirzebruch surfaces are given by $$\Si_k = \{(z, w) \in \CP^2 \times \CP^1 \mid z_0w_0^k + z_1 w_1^k = 0\}$$ for $k \in \N$, when $k$ is even $\Si_k$ is diffeomorphic to $\CP^1 \times \CP^1$, and when $k$ is odd $\Si_k$ is diffeomorphic to $\CP^2$ blown-up at $[0 \pc 0 \pc 1]$. Let $\w_n$ be the Fubini-Study form on $\CP^n$ such that $\int_{\CP^1}\w_n = 1$. Observe that the Poincaré dual of $[\Si_k] \in H_4(\CP^2 \times \CP^1;\Z)$ is $[\O_k] := [\w_2] + k[\w_1]$. Therefore $\P_k = (\CP^2 \times \CP^1, \O_k, J; \Si_k)$ is a polarization of degree $1$. The holomorphic line bundle $\L_k \to \CP^2 \times \CP^1$ determined by $\Si_k$, has an unique (up to scaling) holomorphic section $s_k$ with zero set $\Si_k$. Picking a hermitian metric ${\left\| \cdot \right\|}$ on $\L_k$ gives a function $\r_{k}\co \CP^2 \times \CP^1 \setminus \Si_k \to \R$ where $$\r_{k}(z,w) = -{\left\| s_k(z,w) \right\|}^2 = -\frac{{\left| z_0w_0^k + z_1 w_1^k \right|}^2}{{\left\| z \right\|}^2{\left\| w \right\|}^{2k}}.$$ The skeleton $\Del_k = \Del_{\P_k}$ is defined to be the union of the stable manifolds for $\nabla \r_k$. We have that $$\Del_k = \{[\bar{w}_0^k \pc \bar{w}_1^k \pc 0] \times [w_0 \pc w_1] \mid w \in \CP^1\}$$ which is $2$-dimensional, and therefore the polarization $\P_k$ is subcritical. See [@Bir01 Section 3] for similar examples. To summarize: \[main example\] The polarizations $\P_k= (\CP^2 \times \CP^1, \O_k = \w_2 + k\, \w_1, J; \Si_k)$ given by the Hizebruch surfaces $\Si_k$ are subcritical and the symplectomorphism types of the hypersurfaces are as follows: 1. $(\Si_{2l}, \O_{2l}|_{\Si_{2l}})$ is symplectomorphic to $(\CP^1 \times \CP^1, \w_1 \oplus 3l\, \w_1)$, 2. $(\Si_{2l+1}, \O_{2l+1}|_{\Si_{2l+1}})$ is symplectomorphic to $(\CP^2\, \#\, \overline{\CP^2}, \w)$, where $\w$ is the symplectic form with $\w(L) = 3l+2$ for $L$ the general line and $\w(E) = 3l+1$ for $E$ the exceptional divisor. These two families of rational ruled surfaces have quasi-morphisms and symplectic quasi-states that descend from $(\CP^2 \times \CP^1, \O_k)$, as in Corollary \[cor: subcritical\]. The $\Si_k$ are ruled surfaces, since projection onto the $\CP^1$ factor $\pi_k\co \Si_k \to \CP^1$ gives a $\CP^1$-fiber bundle. Hence by Lalonde and McDuff [@LalMcD96], the symplectomorphism type of $(\Si_k, \O_{k}|_{\Si_k})$ is determined by the cohomology class $[\O_{k}|_{\Si_k}] \in H^2(\Si_k, \Z)$. A basis for $H_2(\Si_k, \Z)$ is given by $$F_k= [\pi_k^{-1}([1\pc b]) ] = [\{z_0 + z_1b^k = 0\} \times [1\pc b]] \quad \mbox{and} \quad D_k = [[0\pc 0 \pc 1] \times \CP^1],$$ where $\O_k(F_k) = 1$ and $\O_k(D_k) = k$, and the intersections are $F_k \cdot F_k = 0$, $F_k \cdot D_k = 1$, and $D_k \cdot D_k = -k$. When $k = 2l$ is even, $(\Si_{2l}, \O_{2l}|_{\Si_{2l}})$ and $(\CP^1 \times \CP^1, \w)$ are symplectomorphic and the homology classes of the spheres are identified with $$[\CP^1 \times pt] = F_{2l} \quad \mbox{and} \quad [pt \times \CP^1] = lF_{2l} + D_{2l}.$$ Therefore $\w([\CP^1 \times pt]) = 1$ and $\w([pt \times \CP^1]) = 3l$, which proves (1). When $k= 2l +1$ is odd, $(\Si_{2l+1}, \O_{2l}|_{\Si_{2l}})$ and $(\CP^2\, \#\, \overline{\CP^2}, \w)$ are symplectomorphic and the homology classes of $L$ and $E$ are identified with $$L = (l+1)F_{2l+1} + D_{2l+1} \quad \mbox{and}\quad E = l F_{2l+1} + D_{2l+1}.$$ Therefore $\w(L) = 3l+2$ and $\w(E) = 3l+1$, which proves (2). ### Subcritical polarizations from algebraic geometry. Biran and Jerby [@BirJer11] have proved that if $M \subset \CP^m$ is a smooth algebraic manifold whose dual variety $M^* \subset \CP^{m*}$ has codimension at least $2$, then a hyperplane section $\Si$ in $M$ gives a subcritical polarization $(M, \Si)$. Examples of such $M$ are the complex Grassmannians $\operatorname{Gr}(2, 2n+1)$ and $\CP^n$-bundles over smooth projective varieties of dimension less than $n$ with linear fibers. Refer to [@Tev03 Section 9.2.C] for other examples. Reduction for critical polarizations. {#critical} ------------------------------------- If $\P = (M^{2n}, \O, J; \Si, \Del_\P)$ is a critical polarization, then $\Del_\P$ has dimension equal to $n$ and is possibly nondisplaceable. Therefore it is not enough that a quasi-morphism have the Calabi property for it to descend, however in light of Proposition \[prop: zero measure\] we can say the following: \[cor: critical\] Let $\P = (M^{2n},\w, J;\Si, \Del_\P)$ be a critical polarization. Spectral quasi-morphisms $\mu\co \CH(M) \to \R$ descend to $\bar\mu\co \CH(\Si) \to \R$ as in Theorem \[thm: descent qm\], provided $\Del_\P$ has a neighborhood $U$ with zero measure $\t_\mu(U) = 0$. \[EP example\] Let $\Si$ be the diagonal in $(\CP^1 \times \CP^1, \w_1 \oplus \w_1)$. This forms a critical polarization and the anti-diagonal $\Del$ is the corresponding skeleton [@Bir01 Section 3.2.1]. The quantum homology $QH_4(\CP^1 \times \CP^1)$ splits into a direct sum of fields, with idempotents $a_{\pm}$ and corresponding spectral quasi-morphisms $\mu_{\pm}$, quasi-states $\z_{\pm}$, and quasi-measures $\tau_{\pm}$. In [@EliPol10 Theorem 1.1] Eliashberg and Polterovich compute that $\Delta$ has a neighborhood $U$ with zero measure $\tau_+(U) = 0$, in particular $\tau_+(K) = 1$ where $K$ is an exotic torus disjoint from $\Del$. Therefore by Corollary \[cor: critical\], $\mu_+$ and $\z_+$ descend to $(\Si, \w_1 \oplus \w_1) = (\CP^1, 2\,\w_1)$. The reduced quasi-state $\bar\z_+$ is the median quasi-state, for it is the unique symplectic quasi-state that vanishes on displaceable sets [@EntPol06 Section 8], but the uniqueness of $\mu([S^2], \cdot)$ as a Calabi quasi-morphism on $Ham(S^2)$ is unknown. Proofs ====== The general set-up {#setup} ------------------ Let $(P, \xi)$ be a closed contact manifold with contact form $\a$ and assume the Reeb vector field $R$ generates a free $S^1$-action on $P$. The quotient $\Si = P/S^1$, has a symplectic form $\s = -\pi_* d\a$ induced by the projection $\pi\co P \to \Si$. From this principal $S^1$-bundle, we can build the *standard symplectic disk bundle $\pi\co (E, \w) \to (\Si, \s)$ modeled on $P$*. The total space $E = P \times_{S^1} \D^2$ has a symplectic form given by $\w = \pi^*\s + d(r^2\a) = -d((1\!-\!r^2)\a)$, where $r$ is the radial coordinate on the open unit disk $\D^2$. Note that $\operatorname{sgrad}r^2 = R$, each fiber has area one (by taking $S^1 = \R/\Z$), and $(\Si, \s)$ embeds symplecticly into $E$ as the zero section. Given a polarization $\P = (M, \O, J; \Si)$, the unit normal bundle $P \to \Si$ is a prequantization space for $(\Si, \O_\Si)$, and hence we can form the standard symplectic disk bundle $(E_\Si, \w_{\operatorname{can}}) \to (\Si, \O_\Si)$. This is the bundle from Biran’s decomposition theorem , compare to [@Bir01 Section 2] and [@Ops11 Section 1]. Returning to the standard symplectic disk bundle $\pi\co (E, \w) \to (\Si, \s)$ modeled on $P$, let $\th\co [0,1] \to \R$ be nonnegative, zero in a neighborhood of one, and such that $\th(r)\co E \to \R$ as a function of the radial coordinate $r$ is smooth. We have an associated linear, order preserving map $$\label{Theta no F} \T\co C^\infty(\Si, \s) \to C^\infty(E, \w) \quad \mbox{defined as} \quad H \mapsto \th \cdot\,\pi^{*}H.$$ By integrating over the fiber, one can check that $$\label{fiber} -\int_0^1 \th'(r)(1-r^2)^n\,dr \int_\Si H\, \s^{n-1} = \int_E \T(H)\, \w^n,$$ so $\T$ preserves the property of a function having zero mean and hence induces a map $\T\co \P\H(\Si) \to \P\H(E)$. At times we will be interested in the maps $\T_\e$ which are induced by $\th_\e(r): [0,1] \to \R$, which smoothly interpolates between $1-r^2$ when $r \leq 1-\e$ and zero when $r \geq 1-\tfrac{\e}{2}$. For ease of notation and future applicability, will prove our lemmas in this general setting. Note that the $\T$ from in Theorems \[thm: descent qm\] and \[thm: descent\] is just $\T$ from combined with the use of the symplectomorphism $F_\P$ from to view $(E_\Si, \w_{\operatorname{can}}) = (E, \w)$ as an open subset of $M$. So our lemmas will be applicable to the setting of the theorems. An overview of the proofs of Theorem \[thm: descent qm\] and \[thm: descent\] ----------------------------------------------------------------------------- Before diving into the details of proving Theorem \[thm: descent qm\] and \[thm: descent\], let’s step back and identify the main issue. We have a homogeneous quasi-morphism $\mu\co \CH(E) \to \R$ and a map $\T\co \PH(\Si) \to \PH(E)$, and we want to form a quasi-morphism $\hat\mu\co \CH(\Si) \to \R$. Our situation can be summarized by the diagram: $$\xymatrix{ \PH(\Si) \ar[d]_{p_\Si} \ar[r]^{\T} & \PH(E) \ar[d] \ar[dr]^{\mu} \\ \CH(\Si) \ar@{-->}@/_1.3pc/[rr]_{\hat\mu} & \CH(E) \ar[r]^{\quad\mu} & \R}$$ If $\T\co \H(\Si) \to \H(E)$ preserved the Poisson brackets, then we would be done for $\T\co \PH(\Si) \to \PH(E)$ would be a homomorphism and would descend to a homomorphism $\widehat{\T}\co \CH(\Si) \to \CH(E)$, which we could use to pullback $\mu$ to a quasi-morphism on $\CH(\Si)$. Unfortunately $\T$ does not preserve the Poisson brackets, but since we only want a quasi-morphism, the full strength of a homomorphism is not necessary. In Section \[group\] we will formulate when a homogeneous quasi-morphism $\mu\co G \to \R$ can be pulled back by a map $H \to G$ (Lemma \[lem:pullback\]), and when $\mu$ will descend along a quotient $G \to H$ (Lemma \[lem:pushforward\]). We will then show that $\mu\co \PH(E) \to \R$ pulls back along $\T_\e$ to a homogeneous quasi-morphism $\T_\e^*\mu\co \PH(\Si) \to \R$, which then descends along $p_\Si$ to a homogeneous quasi-morphism $\hat\mu\co \CH(\Si) \to \R$. The technical heart of these two steps occupies Lemmas \[lem: sgrad\] and \[lem: homotopy\], from which the other parts of Theorem \[thm: descent qm\] and Theorem \[thm: descent\] follow. Two group theory lemmas {#group} ----------------------- For these two lemmas, $G$ and $H$ will be groups and $\mu\co G \to \R$ will be a homogeneous quasi-morphism with defect $D(\mu)$. A map $\vp\co H \to G$ is *$\mu$-quasi-homomorphism* if there is some $D(\vp, \mu)$ such that $${\left| \mu{\left(\vp(h_1h_2)^{-1}\vp(h_1)\vp(h_2)\right)} \right|} \leq D(\vp, \mu) \quad\mbox{for all $h_1, h_2 \in H$}.$$ $\vp$ is *$\mu$-homogeneous* if there is some $C$ such that $${\left| \mu{\left(\vp(h^n)^{-1}\vp(h)^n\right)} \right|} \leq C \quad \mbox{for all $n \in \Z$ and $h \in H$}.$$ \[lem:pullback\] If $\vp\co H \to G$ is a $\mu$-quasi-homomorphism, then $\vp$ pulls $\mu$ back to $\vp^*\mu \co H \to \R$, a quasi-morphism on $H$, with defect bounded above by $D(\vp,\mu) + 2D(\mu)$. If $\vp$ is also $\mu$-homogeneous, then $\vp^*\mu$ is a homogeneous quasi-morphism. We have that $$\begin{aligned} {\left| \mu(\vp(h_1h_2)) \!-\! \mu(\vp(h_1)) \!-\! \mu(\vp(h_2)) \right|} &\leq {\left| -\mu(\vp(h_1h_2)) \!+\! \mu(\vp(h_1)\vp(h_2)) \right|} + D(\mu)\\ &\leq {\left| \mu{\left(\vp(h_1h_2)^{-1}\vp(h_1)\vp(h_2)\right)} \right|} + 2 D(\mu)\\ &\leq D(\vp, \mu) + 2D(\mu), \end{aligned}$$ so $\vp^*\mu$ is a quasi-morphism. If $\vp$ is $\mu$-homogeneous, then $$\begin{aligned} {\left| \mu(\vp(h^n)) - \mu(\vp(h)^n) \right|} &= {\left| \mu(\vp(h^n)^{-1}) + \mu(\vp(h)^n) \right|}\\ &\leq {\left| \mu{\left(\vp(h^n)^{-1}\vp(h)^n\right)} \right|} + D(\mu) \leq C + D(\mu). \end{aligned}$$ Dividing through by $n$ and taking the limit, we see that $\vp^*\mu$ is its own homogenization. \[lem:pushforward\] Let $\vp\co G \to H$ be a surjective homomorphism such that $\mu$ is bounded on the kernel of $\vp$. Then $\vp$ pushes $\mu$ forward to a homogeneous quasi-morphism $$\vp_*\mu\co H \to \R \quad \mbox{where} \quad (\vp_*\mu)(h) = \mu(g) \quad \mbox{for any $g \in \vp^{-1}(h)$},$$ with defect $D(\mu)$. The statement is clear provided that $\vp_*\mu$ is well defined. Let $B \geq 0$ be such that ${\left| \mu(\ker \vp) \right|} \leq B$. Now if $g_1$ and $g_2$ both map to $h \in H$, then $g_1^n$ and $g_2^n$ differ by an element of the kernel, so $$n {\left| \mu(g_1) - \mu(g_2) \right|} = {\left| \mu(g_1^n) - \mu(g_2^n) \right|} \leq B + D(\mu).$$ Now take the limit as $n$ goes to infinity, to get $\mu(g_1) = \mu(g_2)$. The proofs of Theorem \[thm: descent qm\] and Theorem \[thm: descent\] {#proofs} ---------------------------------------------------------------------- We will need the following four lemmas, which are proved in Section \[lemma proofs\]. \[lem: sgrad\] The map $\T\co C^\infty(\Si, \s) \to C^\infty(E, \w)$ preserves the vanishing of Poisson brackets, namely for $H,K \in C^\infty(\Si)$ we have that $$\label{poisson} \{\T(H),\T(K)\}_{\w}= \th(r)^2\{\pi^*H, \pi^*K\}_\w = \frac{\th(r)}{1-r^2}\, \T(\{H,K\}_{\s}).$$ If $\{f_t\} \in \PH(\Si)$ and $\{{\widetilde{f}}_t\} \in \PH(E)$ are generated by $F \in \P\H(\Si)$ and $\T_\e(F) \in \P\H(E)$, then $$\label{commute} \mbox{$f_t \circ \pi = \pi \circ {\widetilde{f}}_t$ when $r \leq 1-\e$.}$$ The term measuring the failure of $\T_\e\co \P\H(\Si) \to \P\H(E)$ to be a homomorphism $$\label{failure} \overline{\T_\e(F\#G)}\,\#\,(\T_\e(F)\#\T_\e(G)) \co E \times [0,1] \to \R$$ vanishes at points in $E$ with $r \leq 1-\e$. This also holds for larger products as well, in particular for $\overline{\T_\e(F^{\#k})}\,\#\,(\T_\e(F)^{\#k})$. \[lem: displace\] If $X\subset \Si$ is displaceable, then ${\widetilde{X}}_a = \pi^{-1}(X) \cap \{r \leq a\} \subset E$ is displaceable if $a < 1$. \[lem: homotopy\] Suppose that $F \in \P\H(\Si)$ generates a loop $f$, which is a null homotopic $[f] = {\boldsymbol{\mathbbm{1}}}$ in $\CH(\Si)$, and let $\T_\e(F) \in \P\H(E)$ generate the path ${\widetilde{f}}$. Then as an element of $\CH(E)$, $[{\widetilde{f}}] = [\eta]$ where $\eta$ is generated by a normalized Hamiltonian on $E$ that vanishes when $r \leq 1-\e$. \[lem: scale\] Let $\mu\co \CH(M) \to \R$ be a stable homogeneous quasi-morphism and let $H \in C^\infty(M)$ be a normalized Hamiltonian. For any smooth function $\l\co [0,1] \to \R$, $$\mu(\p_{\l H}) = {\left(\int_0^1\l(t) dt\right)}\mu(\p_H)$$ where $\l H \in \P\H(M)$ is a time dependent normalized Hamiltonian. ### The proof of Theorem \[thm: descent qm\] By picking $\e_0$ small, we can take the neighborhood $U$ of $\Del_\P$ to be the complement of $F_\P(r \leq 1-\e_0) \subset M$, and we will work with $\e < \e_0$. If $F, G \in \P\H(\Si)$ are normalized Hamiltonians, then by in Lemma \[lem: sgrad\] $$\label{ham1} \overline{\T_\e(F \# G)}\,\#\,(\T_\e(F) \# \T_\e(G)) \quad \mbox{and} \quad \overline{\T_\e(F^{\#k})}\,\#\, \T_\e(F)^{\#k}$$ are normalized and supported on $U$. Since $\mu$ restricted to $\CH_U(M)$ is the Calabi homomorphism, $\mu$ vanishes on since they are normalized. Therefore by Lemma \[lem:pullback\], $\T_\e$ pulls back $\mu$ to a homogeneous quasi-morphism $\T_\e^*\mu$ on $\P\H(\Si)$ with defect at most $2D(\mu)$. It is independent of $\e$ since if $\e', \e < \e_0$, then $\mu$ vanishes on $\overline{\T_\e(F)}\, \#\, \T_{\e'}(F)$, for it is normalized and supported on $U$. Hence $\T_\e^*\mu = \T_{\e'}^*\mu$, for they are homogeneous quasi-morphism a bounded distance apart. If $F \in \P\H(\Si)$ generates a null-homotopic element in $\CH(\Si)$, then by Lemma \[lem: homotopy\] $\T_\e(F)$ generates an element $\eta \in \CH(M)$, which can also be generated by a normalized Hamiltonian in $\P\H(M)$ that is supported in $U$. Therefore by the Calabi property again, $\mu(\eta) = 0$, and hence the quasi-morphism $\T_\e^*\mu$ vanishes on the kernel of $\P\H(\Si) \to \CH(\Si)$. Therefore by Lemma \[lem:pushforward\], $\T_\e^*\mu$ descends to $\CH(\Si)$ as the homogeneous quasi-morphism $\hat\mu$ in with defect at most $2D(\mu)$. Let $\mu$ be stable and have the Calabi property, and recall that $\bar\mu := \z(\th_\e)^{-1}\hat\mu$. Observe that for normalized functions $H, K\co \Si \to \R$, $$\min_{\Si} (H - K) = \min_{M} (\T_\e(H) - \T_\e(K))$$ and likewise for $\max$. Hence the stability of $\bar\mu$ will follow from the stability of $\mu$. The normalization constant $\z(\th_\e)^{-1}$ is independent of $\e$, for small $\e$, because $\th_\e$ and $\th_{\e'}$ Poisson commute and hence $\z(\th_\e) - \z(\th_{\e'}) = \z(\th_\e - \th_{\e'}) = 0$, for the difference is supported on the set $U$ where $\mu$ restricts to the Calabi homomorphism. To prove the Calabi property for $\bar\mu$, let $F\co \Si \times [0,1] \to \R$ be supported on a displaceable set $V \subset \Si$ and let’s denote $\operatorname{Vol}= \operatorname{Vol}(M) = \operatorname{Vol}(\Si)$. Consider the normalization terms for $F$ and $\T_\e(F)$: $$\l(t) = \operatorname{Vol}^{-1} \int_\Si F_t\, \O_{\Si}^{n-1} \quad\mbox{and}\quad \eta(t) = \operatorname{Vol}^{-1} \int_M \T_\e(F_t)\, \O^n = C_\e \l(t),$$ where the last equality uses and the notation $C_\e = \operatorname{Vol}^{-1}\int_M \th_\e\, \O^n$. Since $\T_\e(\l) - \eta = \l\cdot(\th_\e - C_\e)$, by Lemma \[lem: scale\] we have that $$\mu(\p_{\T_\e(\l) - \eta}) = {\left(\int_0^1 \l(t)\, dt\right)} \mu(\p_{\th_\e - C_\e}) = \operatorname{Cal}_V(F)\, \frac{\mu(\p_{\th_\e - C_\e})}{\operatorname{Vol}}.$$ By Lemma \[lem: displace\] the support of $\T_\e(F)$ is displaceable, so therefore $$\mu(\p_{\T_\e(F) - \eta}) = \int_0^1\int_M \T_\e(F_t)\, \O^n dt = C_\e \operatorname{Cal}_V(F) = \frac{\int_M \th_\e\, \O^n}{\operatorname{Vol}}\, \operatorname{Cal}_V(F)$$ Bringing this all together and using that $\T_\e(F)$ and $\T_\e(\l)$ commute gives $$\begin{aligned} \bar\mu(\p_{F-\l}) &= \z(\th_\e)^{-1} {\left(\mu(\p_{\T_\e(F-\l)}) + \mu(\p_{\T_\e(\l) - \eta}) - \mu(\p_{\T_\e(\l) - \eta})\right)}\\ &= \z(\th_\e)^{-1}{\left(\mu(\p_{\T_\e(F) - \eta}) - \mu(\p_{\T_\e(\l) - \eta})\right)}\\ &= \operatorname{Cal}_V(F)\, \z(\th_\e)^{-1} \frac{\int_M \th_\e\, \O^n - \mu(\p_{\th_\e - C_\e})}{\operatorname{Vol}} = \operatorname{Cal}_V(F), \end{aligned}$$ as desired. ### The proof of Theorem \[thm: descent\] The monotonicity and normalization of $\bar{\z}_\th$ are immediate. As for quasi-linearity, if $F, G \in C^\infty(\Si)$ Poisson commute, then by we have that $\T(F), \T(G) \in C^\infty(M)$ Poisson commute as well. Therefore $$\bar{\z}_\th(F + G) = \frac{\z(\T(F)+\T(G))}{\z(\th)} = \frac{\z(\T(F)) + \z(\T(G))}{\z(\th)} = \bar{\z}_\th(F) + \bar{\z}_\th(G),$$ so we have that $\bar{\z}_\r$ is infact a symplectic quasi-state. Suppose that $\z$ vanishes on functions with displaceable support. If $H \in C^\infty(\Si)$ has displaceable support $X \subset \Si$, then by Lemma \[lem: displace\], $\T(H)\co M \to \R$ does as well and hence $\bar\z_{\th}(H) = \z(\T(H)) = 0$. Suppose that $\z\co C^\infty(M) \to \R$ is $Ham(M)$ invariant. Let $\{f_t\}$ be a Hamiltonian isotopy generated by $F \in \P\H(\Si)$. Let $\e$ be small enough so that $\th(r)$ vanishes when $r > 1- \e$, and let $\T_\e(F) \in \PH(M)$ generate the Hamiltonian isotopy $\{{\widetilde{f}}_t\}$. On the image of $F_\P$ in $M$, by we have that $f_t \circ \pi \circ F_\P^{-1} = \pi \circ F_\P^{-1} \circ {\widetilde{f}}_t$, and hence for $H \in C^\infty(\Si)$ we have that $$\T(H \circ f_t)= \th \cdot (H \circ f_t \circ \pi^* \circ F_\P^{-1}) = (\th \cdot (H \circ \pi^* \circ F_\P^{-1}))\circ {\widetilde{f}}_t = \T(H) \circ {\widetilde{f}}_t.$$ Therefore $\bar\z_\th$ is $Ham(\Si)$ invariant, for $$\bar\z_\th(H \circ f_t) = \frac{\z(\T(H \circ f_t))}{\z(\T(1))} = \frac{\z(\T(H) \circ {\widetilde{f}}_t)}{\z(\T(1))} = \bar\z_\th(H).$$ Finally, for the inequality suppose that $\z$ satisfies it with $C(\z)$. Then by , for $H, K \in C^\infty(\Si)$ we have that $$\begin{aligned} \Pi_{\bar{\z}_\th}(H,K) &= \z(\th)^{-1}\, \Pi_\z(\T(H),\T(K)) \leq \frac{C(\z)}{\z(\th)} \sqrt{{\left\| \{\T(H),\T(K)\} \right\|}}\\ &\leq \frac{C(\z)}{\z(\th)} {\left\| \frac{\th(r)}{\sqrt{1-r^2}} \right\|} \sqrt{{\left\| \{H,K\} \right\|}}. \end{aligned}$$ So therefore $C(\bar\z_\th) \leq \frac{C(\z)}{\z(\th)} {\left\| \tfrac{\th(r)}{\sqrt{1-r^2}} \right\|}$. The proofs of Lemmas \[lem: sgrad\], \[lem: displace\], \[lem: homotopy\], and \[lem: scale\] {#lemma proofs} --------------------------------------------------------------------------------------------- We will start by computing $\pi_*(\operatorname{sgrad}\pi^*H)$. For a vector $X \in T_{(p,r)}E$ $$\begin{aligned} -dH(\pi_*X) &= \w(\operatorname{sgrad}\pi^*H, X)\\ &= (\pi^*\s + r^2 d\a)(\operatorname{sgrad}\pi^*H, X) &\mbox{by $\i_{\operatorname{sgrad}\pi^*H}(dr \wedge \a) = 0$}\\ &= (1 - r^2)\, \s(\pi_*(\operatorname{sgrad}\pi^*H), \pi_*X) &\mbox{by $\s = -\pi_*d\a$} \end{aligned}$$ so we conclude that $\pi_*(\operatorname{sgrad}\pi^*H) = \frac{1}{1-r^2}\,\operatorname{sgrad}H$. Therefore $$\label{push sgrad theta} \pi_*(\operatorname{sgrad}\T(H)) = \th(r)\,\pi_*(\operatorname{sgrad}\pi^*H) = \frac{\th(r)}{1-r^2}\,\operatorname{sgrad}H.$$ Now since $\pi^*H$ and $\th(r)$ Poisson commute, the computation for is the following: $$\begin{aligned} \{\T(H), \T(K)\}_\w &= \th(r)^2\{\pi^*H, \pi^*K\}_\w = \th(r)^2\, dH(\pi_*\operatorname{sgrad}\pi^*K)\\ &= \frac{\th(r)^2}{1-r^2}\, \pi^*(\{H,K\}_\s) = \frac{\th(r)}{1-r^2}\, \T(\{H,K\}_{\s}). \end{aligned}$$ Going to the path space side, using we see that $\pi\circ {\widetilde{f}}_t\co E \to \Si$ satisfies the differential equation $$\d_t(\pi \circ {\widetilde{f}}_t) = {\left(\frac{\th_\e(r)}{1-r^2} \operatorname{sgrad}F_t\right)}_{\pi \circ {\widetilde{f}}_t} \quad \mbox{with initial condition } \pi \circ {\widetilde{f}}_0 = \pi.$$ Staring at a point with $\th_\e(r) = 1-r^2$, then $f_t \circ \pi$ also satisfies this equation, so follows by the uniqueness of solutions to ODEs. Denoting by ${\widetilde{(fg)}}_t$ the Hamiltonian path generated by $\T_\e(F \# G)$ we have that $$\begin{aligned} \overline{\T_\e(F\#G)}\,\#\,(\T_\e(F)\#\T_\e(G)) &= (- \T_\e(F\#G) + \T_\e(F)\#\T_\e(G))\circ {\widetilde{(fg)}}_t\\ &= \th_\e(r) \,{\left(-G \circ f_t^{-1} \circ \pi + G \circ \pi \circ {\widetilde{f}}_t^{-1}\right)} \circ {\widetilde{(fg)}}_t, \end{aligned}$$ which vanishes when $r \leq 1-\e$ since then we can use that $f_t^{-1} \circ \pi = \pi \circ {\widetilde{f}}^{-1}_t$. The proof for larger products is the same. Let $F \in \P\H(\Si)$ generate the isotopy $f$ which displaces $X$. Pick $\e > 0$ so that $1-\e > a$ and let ${\widetilde{f}}$ be the isotopy generated by $\T_\e(F) \in \P\H(E)$. It follows from that ${\widetilde{f}}$ displaces ${\widetilde{X}}_a$. Let $\vp_t^s$ be a homotopy of loops in $Ham(\Si)$ based at ${\boldsymbol{\mathbbm{1}}}$, between the loop $\{\vp^0_t = f_t\}_{t\in [0,1]}$ and the constant loop $\{\vp^1_t = {\boldsymbol{\mathbbm{1}}}\}_{t\in [0,1]}$. For $s$ fixed, let $F_t^s\co \Si \to \R$ be the Hamiltonian in $\P\H(\Si)$ generating the Hamiltonian loop $\{\vp^s_t\}_{t\in [0,1]}$ in $Ham(\Si)$, via $$\d_t \vp_t^s = (\operatorname{sgrad}F_t^s)_{\vp_t^s} \quad \mbox{with}\quad \vp_0^s = {\boldsymbol{\mathbbm{1}}}.$$ Note that $F_t^0 = F_t$ and $F_t^1 = 0$. While for $t$ fixed, let $G_t^s\co \Si \to \R$ be the Hamiltonian in $\P\H(\Si)$ generating the Hamiltonian path $\{\vp_t^s\}_{s \in [0,1]}$ in $Ham(\Si)$, via $$\d_s \vp_t^s = (\operatorname{sgrad}G_t^s)_{\vp_t^s} \quad\mbox{with}\quad \vp_t^0 = f_t.$$ They are related by the equation [@Ban78 Proposition I.1.1] $$\label{dhf1} \d_s F_t^s = \d_t G_t^s + \{F_t^s, G_t^s\}.$$ Fixing $s$, the Hamiltonian $\T_\e(F^s_t)\co E \to \R$ in $\P\H(E)$ will generate a Hamiltonian path $\{\psi^s_t\}_{t\in [0,1]}$ in $Ham(E)$, via $$\d_t \psi_t^s = (\operatorname{sgrad}\T_\e(F_t^s))_{\psi_t^s} \quad\mbox{with} \quad \psi_0^s = {\boldsymbol{\mathbbm{1}}}.$$ As $s$ varies, $\psi_s^t$ will be a homotopy of Hamiltonian paths in $Ham(E)$, between the paths $\{\psi_t^0 = {\widetilde{f}}_t\}_{t\in [0,1]}$ and $\{\psi_t^1 = {\boldsymbol{\mathbbm{1}}}\}_{t\in [0,1]}$. However this will not be a homotopy of loops, since in particular $\{{\widetilde{f}}_t\}_{t\in [0,1]}$ may not be a loop. Letting $t$ be fixed, let $H_t^s\co E \to \R$ be the Hamiltonian in $\P\H(E)$ generating the Hamiltonian path $\{\psi_t^s\}_{s \in [0,1]}$ in $Ham(E)$, via $$\d_s \psi_t^s = (\operatorname{sgrad}H_t^s)_{\psi_t^s} \quad \mbox{with} \quad \psi_t^0 = {\widetilde{f}}_t.$$ Just as in , we have the relation $$\label{dhf2} \d_s \T_\e(F_t^s) = \d_t H_t^s + \{\T_\e(F_t^s), H_t^s\}.$$ Define the path $\{\eta_u = (\psi_1^u)^{-1}\psi_1^0\}_{u \in [0,1]}$ in $Ham(E)$, which is a path from $\eta_0={\boldsymbol{\mathbbm{1}}}$ to $\eta_1 = \psi_1^0$, using that $\psi_t^1 = {\boldsymbol{\mathbbm{1}}}$. Observe that $\Psi_t^s = \psi_t^s\eta_{st}$ is a homotopy of paths in $Ham(E)$, between the paths $$\{\Psi^0_t = \psi^0_t = \tilde{f}_t\}_{t \in [0,1]} \quad \mbox{and} \quad \{\Psi_t^1 = \eta_t\}_{t \in [0,1]}.$$ Since $\Psi$ is a homotopy of paths with fixed endpoints, $$\Psi^s_0 = {\boldsymbol{\mathbbm{1}}}\quad \mbox{and} \quad \Psi_1^s = \psi_1^s \eta_s = \psi_1^0,$$ we have proved that $[{\widetilde{f}}] = [\eta]$ in $\CH(E)$. The path $\eta$ is generated by the normalized Hamiltonian $(-H_1^u \circ \psi_1^u)$, which we claim vanish at points with $r \leq 1-\e$. It suffices to prove this for $H_1^u$ since $\psi_t^s$ preserves the level sets of $r$. Applying $\T_\e$ to and using of Lemma \[lem: sgrad\] gives that $$\label{dhf3} \d_s \T_\e(F_t^s) = \d_t \T_\e(G_t^s) + \frac{(1-r^2)}{\th_\e(r)}\, \{\T_\e(F_t^s), \T_\e(G_t^s)\}.$$ When $r \leq 1-\e$, we have that $\th_\e(r) = 1-r^2$ and so the differential equations and agree. Therefore by the method of characteristics for PDE’s, $H_t^s = \T_\e(G_t^s)$ when $r \leq 1 - \e$. The fact that $\vp$ is a homotopy of paths with fixed endpoints means that $G_1^s = 0$, and therefore $H_1^s = 0$ when $r \leq 1-\e$. If $L(t) = \int_0^t \l(s)ds$ and $h_t$ is the flow generated by $\operatorname{sgrad}(H)$, then the Hamiltonian $\l H$ generates the Hamiltonian isotopy $\p_{\l H} = \{h_{L(t)}\}_{t\in [0,1]}$. By a time reparameterization, in $\CH(M)$ we have that $[\{h_{L(t)}\}] = [\{h_{tL(1)}\}]$ and hence $\mu(\p_{\l H}) = \mu(\p_{L(1)H})$. If $m$ is an integer, then $\mu(\p_{mH}) = \mu(\p_H^m) = m\mu(\p_H)$ since $\mu$ is homogeneous and hence $\mu(\p_{a H}) = a \mu(\p_{H})$ when $a$ is rational. Since $\mu$ is stable, this extends to any real scalar, so in particular $\mu(\p_{L(1)H}) = L(1) \mu(\p_{H})$. The proof of Theorem \[thm: stably\] {#proofs2} ------------------------------------ Since $X \times S^1$ is displaceable in $M \times T^*S^1$, it is also displaceable in $M \times S^1 \times [-R, R]$ where the last coordinate is the vertical direction in $T^*S^1$. Capping $S^1 \times [-R, R]$ off on the top and bottom with half spheres creates an sphere. Therefore by Moser we may assume that $X \times S^1 \subset M \times (S^2, \O)$ is displaceable, where $(S^2, \Omega)$ is a round sphere in $\R^3$ with coordinates $(x,y,z)$, $\O$ is the induced area form, and $S^1$ is the equator $\{z=0\} \subset S^2$. Let $U$ and $V = \{{\left| z \right|} < 2\e\}$ be open neighborhoods of $X$ and $S^1$ respectively such that $U \times V \subset M \times S^2$ is still displaceable. Let $\rho\co M \to \R$ be a cut-off function with support in $U$ that is a constant $1$ near $X$. On $S^2$, let $V_1 = \{z> \e\}$ and $V_2 = \{z< -\e\}$, so $\{V=V_0, V_1, V_2\}$ is an open cover of $S^2$ and let $\phi_i(z)$ be a Poisson commuting subordinate partition of unity. Note that each $U \times V_i \subset M \times S^2$ is displaceable. Let $\pi_1\co M\times S^2 \to M$ and $\pi_2\co M \times S^2 \to S^2$ be the projections and let $\z(a\otimes[S^2], \cdot)$ be the partial symplectic quasi-state on $M \times S^2$ associated to the idempotent $a \otimes [S^2] \in QH_{2n+2}(M \times S^2)$, see [@EntPol09RS Section 3.5]. Since the $\phi_i$ pullback to a Poisson commuting partition on unity on $M \times S^2$, $$\z(a \otimes [S^2], \pi_1^* \rho) = \z\left(a \otimes [S^2],\,\sum_{i=0}^2 \pi_1^*\rho \cdot \pi_2^* \phi_i\right) = 0.$$ The above vanishes by the partial additivity and vanishing property [@EntPol09RS Theorem 3.6] of partial symplectic quasi-states, using that the $\pi_1^*\rho \cdot \pi_2^* \phi_i$ Poisson commute and have displaceable support. Finally by the product formula for spectral invariants [@EntPol09RS Theorem 5.1], we have that $$\z(a, \rho) = \z(a \otimes [S^2], \pi_1^*\rho) = 0.$$ It now follows that from the definition of $\tau(a)$ that $\tau(a, X) = 0$. Questions ========= We will end this paper with two questions regarding the process of passing from a quasi-morphism $\mu: \CH(M) \to \R$ to its reduction $\bar\mu: \CH(\Si) \to \R$ as described in Theorem \[thm: descent qm\]. 1. If $\mu$ descends from the universal cover to $\mu: Ham(M) \to \R$, does its reduction also descend to $\bar\mu: Ham(\Si) \to \R$? 2. If $\mu = \mu(a, \cdot): \CH(M) \to \R$ is a spectral quasi-morphism for some idempotent $a \in QH_{2n}(M)$, then does its reduction $\bar\mu = \mu(\bar{a}, \cdot)$ come for some idempotent $\bar{a} \in QH_{2n-2}(\Si)$? The trouble with proving 1) is that $\T_{\e}: \PH(M) \to \PH(\Si)$ does not preserve loops due to the discrepancy between $\th_{\e}(r)$ and $1-r^{2}$. For 2) the trouble is that there is no natural candidate for $\bar{a}$ since there is no known natural map $QH_{2n}(M) \to QH_{2n-2}(\Si)$ preserving the ring structure. [^1]: This work is partially supported by the NSF-grant DMS 1006610. | Mid | [
0.593582887700534,
27.75,
19
] |
Sophie’s journey: Three countries, three stories, same abuse Emel Coşkun18/07/2016 Female migrant labourers face discrimination and marginalisation all over the world. Sophie’s stories from Dubai, Turkey and Saudi Arabia can tell us a lot about their experiences. Sophie was looking tired when I met her in Kampala in January 2016. We had last met the summer before in Turkey when she told me her employer had refused to pay her wages. I offered to help her make an official complaint but she just shrugged and said “Forget about it – nothing will come of it except for trouble.” In Uganda, she told me how she had finally made it home but as we talked I realised that her trouble in Turkey was only one stage of a journey of exploitation that had also taken her to Dubai and Saudi Arabia. Sophie’s depressingly similar experience in three countries over five years shows how gender and international migration regimes can conspire against migrant workers. From house wife to breadwinner Sophie’s life had always been a tough one. At 17 she married a man who promised to support her to study but after three children and still no education, Sophie had enough. She left her children and funded herself to College level where she qualified as a masseuse. “This therapy job fed my children and took them through school” she says. But things were still hard and when a client in Kampala offered her a job in Dubai, she took the chance: The client paid for my visa, ticket everything… but when I reached Dubai it was a different story. I understood then that she was a prostitute. She picked me from the airport to one of the rooms she rented to a group of prostitutes. I asked how much I would have to pay back to her because I had to send money to my children. She told me that I’d work for her for 2 months… I was a housemaid. I was always waiting for them, cooking nice African food and ready to open the door if the Police were chasing them. We started selling food and then, I started doing massage in the house. I was making a lot of money for her, may be US$10,000 for two months. I was doing everything… house work, massage, food. But Sophie’s employer wanted more and tried to make her to sell sex. She refused – and that’s when the mistreatment began. “I tried to find another job for myself in Dubai but my visa was finished and I couldn’t find a thing”. So after three months she left Dubai having earned just enough to cover the fine for overstaying her visa. “I don’t mind working hard if I am getting money” Back in Uganda and back where she’d started, Sophie still faced the pressure of needing to support her children and she still lacked the money necessary to start her business. So she decided to migrate again. This time, her destination was Turkey, where she’d heard that one could earn as much as $350 a month. “I don’t mind working hard if I’m getting money”, she said. She decided not to use an intermediary or an agent and raised more than $2000 to get to Turkey on her own. People at the airport in Istanbul ask for $200 to look after you for the first weeks. They’re mostly Ugandans and it was a street woman who received me. If they find you a job, you give them $50 and you have to pay for the transportation card, food, etc. Everything costs money and work is scarce. Sophie’s first job was in a textile factory. She worked for two months, nine hours a day, six days a week. It was hard. “Even staying in the toilet for too long was not allowed”, she said. She was paid for only two weeks of her work: “they said we shall pay…we shall pay…but they never did”. Employers take advantage of newcomers who know nothing about working in Turkey, don’t speak Turkish, and don’t have work permits. When I met Sophie, she had been in Istanbul for couple of months and in her second factory. Sophie was getting paid her but she only had a few hours work a week. She decided to find another job and, through friends, she found a massage job in the tourist city of Antalya. Again, she worked for two months without pay. He did not pay at all because he knows that I was far from Istanbul and would not go anywhere. I had no work permit so I had no rights. After 2 months I left to work with another woman but I had already made up my mind to go home. I was going back to Uganda in 3 months. I thought let me try to work here and make some money so I worked with them. I made 250 euro, 100 dollar and 200 lira. Getting a work permit in Turkey is a long and difficult bureaucratic process. Employers must apply for migrant workers before they arrive in Turkey or after they are granted a six months resident visa. The lack of a work permit creates discrimination that suits many employers as undocumented migrant workers are too vulnerable to claim their rights. Black migrants also face discrimination on grounds of race and Sophie faced racism everywhere she worked. She explains: The Russian guests don’t like blacks, but Germans are OK and Danish too. All the staff expect you to clean up the place although we were all colleagues, they were masseuses and so was I. I had the same pay as women workers from Russia and Georgia, but as a black worker I had to do the cleaning. She stayed 11 months in Turkey. She saved nothing more than the exit fine for overstaying her visa, again. The Journey to Saudi Arabia Back in Uganda, Sophie struggled to find a good job, so she decided to take another chance and go to Saudi Arabia as a housemaid. This time she used an agency who didn’t ask for money upfront. All her expenses, except for the medical report to prove she is not HIV+, were paid by the agency. Sophie describes this third migration as like slavery with the agencies, licensed by the Saudi government, as little more than mechanisms for worker exploitation. Sophie explains: Arab families pay a lot of money to get a housemaid and they claim the money back if the housemaid doesn’t stay long. As a maid you have no voice, the deal is between the agency and the family. The broker who takes you are doesn’t tell you what is waiting for you because they are just looking at the money. The agency pays you and you have a contract with the agency but it is fake. For example, the contract might say you work for 9 hours per day but it is not true. When she arrived in Riyadh she was kept in a hostel and taught how to clean. She was also given some basic orientation and some Arabic words. The hostels were “like a prison”, with people locked inside. “They give you a profile picture and you are on the computer – they sell you through the computer. They give you a uniform and hijab and then they pass you out”. Sophie worked for a family for 3 weeks before she had to leave: The smaller apartment had 6 or 7 rooms and a big living room. You are mopping the floor, cleaning everything. When you finished working there, they lend you to other family members. They have married sons and daughters and you also work for them. You eat after everyone else and they don’t think if you have eaten anything. People have collapsed – starved and collapsed! You receive $150 but you never know how much the family pays the agency. Sophie felt like a “slave” and lacked for sleep and food. She wanted out but leaving meant breaking her contract. “I told the family and the agency that I wasn’t well. The agency was pushing me to stay and they wanted to just dump me in another hostel”. She managed to get back to Uganda but only by paying the agency off. So again, she made nothing from her trip. Being a female migrant worker Sophie is a strong woman. She was unafraid to tell her story and refused to keep quiet about her experiences. Partly because she went public, the Ugandan government has banned businesses sending maids to Saudi Arabia but, as Sophie says, the trade has just gone underground and Ugandans are now being smuggled across borders. Outlawing the agents has done nothing to alter the fact that poor people seeking to earn money through migration have to risk of precarious, undocumented lifestyles and many forms of exploitation. Sophie had hoped migration would allow her to earn capital to start a business that would support her kids. But she’s worked in three different countries and has been exploited in all of them. Agents, brokers, smugglers and employers have all made money from her labour while she’s never made more than the fine to pay her way out and return home. On the face of it she has ‘failed’ as a migrant but she has been a victim of unequal social relations that separate migrants from citizens and which are continually reinforced by immigration and labour market controls. Saudi Arabia, Turkey and Dubai share migration regimes that maintain an uneven playing field and keep migrants expendable. These regimes create highly precarious migrant groups who can be exploited by various third parties. In all three different countries women like Sophie are abused because of their gender and their race. Migrant women are mostly employed as domestic workers, waitresses, sweatshop workers or prostitutes where local assumptions about their gender and race imagine they belong. Sophie is now working for a hotel in Kampala as a masseuse. She works 10-12 hours per day. She is happy: “I am working really hard but I feel at home.” She still dreams of opening a massage salon in Kampala and she’s still looking for the capital she needs. About the author: Emel Coşkun is an assistant professor of sociology at Düzce University, Turkey. She received her PhD in Migration Studies at the University of Kent in 2013. Her research interests relate to the experiences of women migrants; gendered migration flows; undocumented migration; prostitution and migration politics. Her recent articles focus on sex trafficking and undocumented women migrants in Turkey. | Mid | [
0.5829596412556051,
32.5,
23.25
] |
Introduction ============ In our bodies, the intestinal environment provides a shelter for an enormous number of microorganisms; bacteria comprise up to 60% of the dry mass of feces. Other types of intestinal microorganisms include fungi, archaea and viruses. Compared with that of bacteria, the incidence of pathogenic fungi infections is relatively low; therefore, we might easily undervalue their potential threat to our health. In fact, several studies have identified that the gut fungal microbiota also exert important functions in the gut \[[@goz028-B1], [@goz028-B2]\]. In contrast to the commensal bacterial community, the fungal community is unstable and prone to higher variability in the gut \[[@goz028-B3]\]. Whereas recent studies have found that the fungal microbiota is more stable than originally thought, it is significantly altered in certain intestinal diseases, such as inflammatory bowel disease (IBD) \[[@goz028-B4], [@goz028-B5]\]. In the gastrointestinal tract of healthy individuals, only a small number of fungal species, including *Candida*, *Saccharomyces* and *Aspergillus*, colonize and grow steadily \[[@goz028-B6]\]. Dysbiosis of the gut bacteria has been studied for several decades and this specific term describes a perturbed condition involving alterations in the gut flora that can be both result from and cause various pathologies \[[@goz028-B9]\]. The gastrointestinal tract is the longest visceral cavity in the body, with a length of approximately 16 feet and a massive absorptive area due to its microvilli \[[@goz028-B10]\]. As it is in direct contact with food-derived antigens, the commensal gut microbiota and potential invasive pathogens that enter through the diet, the gut immune system has developed sophisticated mechanisms for sensing pathogenic and non-pathogenic microorganisms and for initiating an immediate response under threats from pathogens \[[@goz028-B1]\]. The gut immune system is necessary for homeostasis maintenance in the gut microenvironment, as it maintains a delicate balance among bacteria, fungi and the gut epithelium. The recognition of fungal pathogens and initiation of the subsequent antifungal immune responses in the gut are mostly mediated by members of the C-type lectin receptor (CLR) family. CLRs play critical roles in gastrointestinal antifungal immunity by serving as pattern-recognition receptors (PRRs). Some key members include Dectin-1, Dectin-2, Dectin-3 and Mincle. These CLRs recognize pathogen-associated molecular patterns (PAMPs) expressed on certain fungal cell walls and induce innate and adaptive immune responses \[[@goz028-B11]\]. IBD is a common and complex intestinal disease that is clearly associated with intestinal dysbiosis \[[@goz028-B12], [@goz028-B13]\]. IBD has two main clinical manifestations: Crohn's disease (CD) and ulcerative colitis (UC) \[[@goz028-B14]\]. Both bacteria and fungi can perturb harmonious host--microbe symbiosis and immune homeostasis. Therefore, CLRs, as innate immune receptors, may play vital roles in IBD progression. There is also a link between cancer and the gut microbiota \[[@goz028-B15]\]. Since cancer susceptibility and severity are associated with inflammation \[[@goz028-B16]\], there may be a relationship between oncogenesis, tumor progression and antifungal immunity \[[@goz028-B17]\] because the antifungal immunity is mainly mediated by CLRs. The central roles that CLRs play in promoting immunity to fungal pathogens are also shared with other PRRs such as Toll-like receptors (TLRs) \[[@goz028-B18]\], suggesting that it may be possible to target some of these receptors, such as Toll-like receptor 1 or Toll-like receptor 9, to induce more efficient immune responses. The collaborations between CLRs and other PRRs might form the basis for new combinative approaches for fighting gastrointestinal disorders and related diseases. Mucosal immunity to the gut microbiota ====================================== The gastrointestinal microbiota plays a critical role in shaping host immunity. In return, the host immune system uses a variety of mechanisms to maintain a harmonious symbiosis \[[@goz028-B21]\]. This process involves the induction of protective immune responses to pathogens as well as the regulatory immunity required for ongoing tolerance of resident microbiota during homeostasis and the immune responses during dysbiosis \[[@goz028-B8], [@goz028-B22]\]. Our knowledge about mucosal immunity to the gastrointestinal microbiota mostly came from work performed under conditions of pathogen infection, while the functions of the immune system and the microbiota during homeostasis or dysbiosis remain unclear. In this section, we discuss the regulatory roles of the gastrointestinal immune system in the maintenance of homeostasis with a focus on the fungal community. Alterations in the intestinal mycobiota may also be associated with the susceptibility and severity of various diseases. Beneficial fungi and pathogenic fungi can be identified in the gastrointestinal tract during homeostasis and dysbiosis. The beneficial fungi are usually defined as harmless commensals living in the gut and the appropriate proportion of the beneficial fungi remains stable to maintain a delicate balance. In contrast, pathogenic fungi are found to be increasing in many diseases and influence the progression of the disease. *Saccharomyces cerevisiae* is thought to be beneficial to gut microbial ecology and has probiotic effects, while, if co-morbidities are present, *S. cerevisiae* may give rise to the increasing incidence of *Saccharomyces* fungemia \[[@goz028-B6]\]. This example suggests that the pathogenic fungi and beneficial ones can interconvert into each other. Many factors could contribute to fungal dysbiosis, including use of antibiotics, diet and genetic factors. Compared with intestinal fungal infections, which are relatively rare, fungal dysbiosis occurs more frequently in the gastrointestinal tract. For example, *Candida albicans* is a well-known commensal in the gut of healthy individuals; however, it can also be pathogenic in the gastrointestinal tract. A shift from a normal commensal state to a dangerous pathogen state can be caused by antibiotic use. One study found that treatment with oral antibiotics can lead to overgrowth of *Candida* populations in the gut of people and mice \[[@goz028-B23]\]. Furthermore, the symbiosis between bacteria and fungi may also affect how the host immune responses control fungal colonization and growth. Mice colonized with *Bacteroides thetaiotamicron* are more resistant to the colonization of *C.* *albicans* than the germ-free mice via production of anti-microbial peptide LL-37 \[[@goz028-B24]\]. PRRs are often activated during infection; however, CLRs also exert their important host-protective functions during dysbiosis. *C. albicans* is a common commensal microbiota in the gut. Dectin-2 knockout mice were more susceptible to *C.* *albicans*, indicating the role of Dectin-2 in recognizing the yeast in its commensal state and in suppressing the overgrowth of this potentially pathogenic microorganism \[[@goz028-B25]\]. Our understanding of normal fungal diversity and alterations therein based on diet or other environmental factors is still at an early stage. Nevertheless, an increasing number of studies suggest that the diversity in the intestinal fungal microbiota is skewed in many diseases such as autism, obesity and asthma \[[@goz028-B26]\]. During homeostasis, commensal fungi exert many of their physiological effects in the gastrointestinal tract by facilitating nutrient absorption and digestion via production of enzymes and vitamins. A study found that *Bacteroides thetaiotaomicron*, a prominent member of the gastrointestinal microbiota, uses yeast mannan as its food source via a selfish mechanism, indicating an additional way in which the fungal microbiota could participate in gastrointestinal physiology \[[@goz028-B27]\]. Since CLRs are mainly expressed on the immune cells in the lamina propria, fungal molecules rarely reach CLRs during homeostasis due to the normal regeneration capability and integrity of the mucosal surface epithelium. Structure and activation of CLRs ================================ CLRs are the main PRRs that mediate the antifungal immune responses in the gut. They can be differentiated from other PRRs by the C-type lectin domain in their extracellular region ([Figure 1](#goz028-F1){ref-type="fig"}). The PAMPs that CLRs recognize are the carbohydrates on the pathogen surface \[[@goz028-B28]\]. The intestinal mucosal barrier is the first line of defense in the protection of our gut against infection \[[@goz028-B4]\]. However, many factors can damage this barrier, which can then allow invasion by intestinal pathogens, such as a lack of or over activation of certain immune responses due to disruption of the gut flora \[[@goz028-B29]\]. Furthermore, microfold cells (M cells) are an important contributing factor because they serve as a portal, through which the pathogen can cross the barrier to induce subsequent immune responses in the intestinal phagocytes of the lamina propria \[[@goz028-B30]\]. During this process, CLRs play various important roles of mediating immune recognition and response. {#goz028-F1} Dectin-1 -------- Dectin-1 is primarily expressed by myeloid cells and serves as the predominant PRR for mediating antifungal immunity in the gut. β-glucans are the major PAMPs present within the fungal cell walls that are recognized by Dectin-1. Dectin-1 expression was first identified in dendritic cells (DCs) and it was later shown to be widely expressed in various cell types, including monocytes, macrophages and neutrophils \[[@goz028-B31]\]. Dectin-1 is a 28-kDa type II membrane protein. An extracellular C-type lectin-like domain is connected by a stalk to a transmembrane region and this stalk contains a special region. The length of the stalk varies to facilitate β-glucan recognition and a longer length induces better binding efficiency. This region is defined as the carbohydrate-recognition domain (CRD). Trp221 and His223 are the two crucial residues for β-glucan binding. The Dectin-1 cytoplasmic tail contains an immunoreceptor tyrosine-based activation motif (ITAM)-like motif, which is involved in the activation of intracellular signaling. In contrast to classical CLRs, Dectin-1 is atypical due to its metal-ion-independent carbohydrate recognition \[[@goz028-B32], [@goz028-B33]\]. Upon ligand binding, Dectin-1 is a tyrosine phosphorylated by Src kinase. This phosphorylation leads to docking and activation of the tandem SH2-domain-containing protein spleen tyrosine kinase (Syk). Through the PKCδ- caspase recruitment domain family, member-9 (CARD9)--B-cell lymophoma-10 (Bcl-10)--mucosa-associated lymphoid tissue lymphoma translocation protein 1 (MALT1) axis, nuclear factor-κB (NF-κB) is finally activated, inducing an intracellular signaling cascade that mediates various cell-specific responses, such as reductions in the levels of various cytokines and chemokines, the respiratory burst and phagocytosis-mediated ligand uptake. The induction of adaptive immune responses, particularly the T helper 1 (Th1) and T helper 17 (Th17) cell responses, is also directed by Dectin-1. Furthermore, Dectin-1 collaborates with other PRRs, such as Toll-like receptor 4 (TLR4) and Dectin-2 \[[@goz028-B34], [@goz028-B35]\]. As discussed above, β-glucan recognition by Dectin-1 is the start of phagocytosis and the initiation of protective immune responses. The immunological potency of β-glucans comes from their molecular mass, spatial stereoscopic structure and degree of branching \[[@goz028-B36]\]. The natural 3D structure of β-glucans depends on these three characteristics. β-Glucan species of longer lengths cannot form a crystal lattice in complex with Dectin-1 \[[@goz028-B37]\]. However, if the length of the β-glucan is too short (less than 10- or 11-mer), the binding will be weak. Furthermore, Dectin-1 oligomerization also increases the affinity for β-glucans. Additionally, whether the binding affinity of Dectin-1 is related to the type of β-glucan remains unclear \[[@goz028-B38]\]. The triple helix conformation of β-glucan enhances its biological functions related to its antitumor activity \[[@goz028-B39]\]. DCs and macrophages are activated and initiate phagocytosis after recognition of β-glucan by Dectin-1. However, if the β-glucan is fragmented, DCs and macrophages cannot be activated \[[@goz028-B40]\]. Dectin-2 -------- Dectin-2 is a C-type lectin-like receptor that is expressed in myeloid cells and it is mainly expressed on DCs, monocytes and macrophages. Unlike Dectin-1, which recognizes β-glucans, Dectin-2 is a PRR that binds mannose and its CRD contains a glutamic acid-proline-asparagine (EPN) motif, which is Ca^2+^-dependent. Unlike Dectin-1, which recruits and activates Syk directly, Dectin-2 interacts with an Fc receptor γ subunit (FcRγ) on the cell surface that contains a Syk-interacting immunotyrosine activation motif. A range of signaling pathways is initiated that ultimately lead to NF-κB activation, which then promotes the production of various cytokines and mediates the subsequent innate and adaptive immune responses. Furthermore, Dectin-2 and other CLRs (including Dectin-3) can functionally collaborate. Dectin-2 and Dectin-3 interact in a heterodimeric form to mediate the defense against fungal infection \[[@goz028-B41], [@goz028-B42]\]. Mincle ------ Mincle is a member of the Dectin-2 family of CLRs and it is mainly expressed on activated macrophages. Mincle was first discovered to be an important player in gastrointestinal anti-microbial immunity based on its ability to bind a variety of PAMPs derived from pathogenic fungal microbiota, including α-mannose, lipidic species and some endogenous self-ligands, such as Sin3A-associated protein 130 (SAP130), which is a product of dead cells. Dectin-2 is a type II transmembrane protein with a short cytoplasmic tail and its extracellular domain is responsible for ligand binding. Like that of Dectin-2, Mincle-dependent signaling also occurs via FcRγ culminating in a pathway involving Syk and the CARD9--Bcl-10--MALT1 complex to activate NF-κB, which then leads to several pro-inflammatory responses and adaptive immune responses against pathogens \[[@goz028-B43], [@goz028-B44]\]. Dectin-3 -------- Unlike the other three receptors mentioned above, the characteristics and functions of Dectin-3 remain obscure. Dectin-3, known also as *Mcl/Clecsf8/Clec4d*, is a myeloid cell-specific CLR family member. The PAMPs it recognizes are mainly mycobacterial trehalose 6, 6-dimycolate (TDM) as well as α-mannans that exist in the fungal cell wall. Unlike Dectin-1, which contains an ITAM in its cytoplasmic region, Dectin-3 has no signaling motif in its cytoplasmic domains; instead, it signals downstream via the ITAM-containing adaptor molecule FcRγ. The CRD of Dectin-3 is Ca2^+^-independent for carbohydrate recognition. Dectin-3 may use a signaling pathway similar to those of the other three CLRs to mediate intestinal antifungal immunity. More studies are required to gain more detailed and precise information about its structure and activation \[[@goz028-B42], [@goz028-B45], [@goz028-B46]\]. CLRs-mediated immune recognition of the gut MICROBIOTA ====================================================== Recognition of invading pathogens by the host is the first step and a key stage in initiating an effective immune response after gastrointestinal fungal infection \[[@goz028-B47]\]. Efficient immune recognition of microorganisms initiates the crosstalk between the host and gut microbiota. And this process in turn helps to maintain gastrointestinal homeostasis via tolerance of the normal commensal microorganisms. CLRs play key roles in immunity to fungal pathogens in the gut microbiota. These receptors bind to almost all the pathogenic fungi found in the gut. The recognition specificity of the CLRs is largely based on different carbohydrate structures expressed by different fungal species as well as various morphological forms of the same organism \[[@goz028-B48]\]. Different fungal PAMPs can be exposed because the morphological forms of the gut fungi are constantly alternating between the yeast and hyphae state \[[@goz028-B49]\]. For example, the cell wall of *C.* *albicans* mainly contains carbohydrate polymers and glycoproteins. The components of the cell wall comprise mainly chitin, β-1, 3- and β-1, 6-glucans and O- and N-linked mannan. In *C.* *albicans*, the β-1, 3- and β-1, 6-glucans are shielded by mannoproteins, while, in the budding yeast and hyphae forms, they are expressed on the surface of the cell wall. In contrast, the cell wall of *Aspergillus* predominantly contains chitin, α-glucans, β-1, 3-glucans and β-1, 4-glucans, and galactomannans \[[@goz028-B50]\]. As shown in [Table 1](#goz028-T1){ref-type="table"}, different CLRs recognize and bind to the various PAMPs expressed on the fungi and mediate the downstream immune responses. Dectin-1, as discussed above, is the primary CLR in the gut that recognizes β-1, 3-glucan chains. Therefore, both *C.* *albicans* and *Aspergillus fumigates* can be recognized by Dectin-1. Dectin-1 can only recognize particulate β-glucans to induce cytokines and initiate internalization of the fungus \[[@goz028-B50]\]. A recent study found that, upon infection, Dectin-1 can modulate interleukin-1 receptor-associated kinase 1 (Irak1) and receptor-interacting protein 2 (Rip2), which are the key adaptor proteins in the TLR and Nod-like receptor signaling pathways, respectively \[[@goz028-B51]\]. This finding indicates that *A. fumigatus* recognition by Dectin-1 may involve interplay with other PRRs in the gut. Human primary myeloid DCs can also interact with *A. fumigates* via Dectin-1 \[[@goz028-B52]\]. As Dectin-2 recognizes mannans, the mycobiota that Dectin-2 recognizes are *C. albicans*, *Candida glabrata* and *Cryptococcus neoformans* \[[@goz028-B53]\]. Dectin-2 can cooperate with Dectin-3 by forming heterodimers to recognize *C.* *albicans* \[[@goz028-B45]\]. Dectin-2 is also involved in *A. fumigatus* recognition and its expression is restricted to macrophages \[[@goz028-B54]\]. A recent study found that Dectin-2 participates in the recognition of *Pneumocystis*, which is a potential gut pathogen \[[@goz028-B55]\]. *Cryptococcus* *neoformans* is a yeast-type opportunistic fungal pathogen that is normally associated with the respiratory system and can damage the central nervous system. Dectin-2 also participates in *C. neoformans* recognition and mediates the antifungal immunity against this pathogen \[[@goz028-B56]\]. Although the binding ligand for Mincle in the gastrointestinal fungi remains unclear, Mincle recognizes *C. albicans* and mediates the macrophage-dependent immune response via TNF-α production \[[@goz028-B57]\]. It is interesting that Mincle recognizes not only fungi, but also *Mycobacterium* via the TDM glycolipid (also known as cord factor) cell-wall component \[[@goz028-B58]\]. ###### C-type lectin receptors and their respective PAMPs and microbiota involved in gastrointestinal immunity CLR Ligand(s) PAMP Microbiota References ------------------------------------------------- ------------------- --------------- ---------------------- ---------------------------------- Dectin-1 β-1, 3-glucan β-1, 3-glucan *Candida albicans* \[[@goz028-B50]\] *Aspergillus fumigatus* Dectin-2 α-mannan α-mannan *Candida albicans* \[[@goz028-B45], [@goz028-B53]\] *Candida glabrata* *Cryptococcus neoformans Aspergillus fumigatus* Dectin-3 α-mannose α-mannose *Candida tropicalis* \[[@goz028-B46], [@goz028-B62]\] mycobacterial trehalose6, 6-dimycolate Mincle Glyceroglycolipid α-mannose *Candida albicans* \[[@goz028-B57]\] mannosyl fatty acids α-mannose CLR, C-type lectin receptor; PAMP, pathogen-associated molecular pattern. Mincle is also associated with recognition of other bacteria such as *Listeria monocytogenes*, *Klebsiella pneumonia*, *Streptococcus pneumonia*, *Pneumocystis pneumonia* and *Escherichia coli*, and all of these microorganisms are thought to be associated with CD \[[@goz028-B59]\]. These clues indicate that we may pay more attention to the possible relationship between Mincle and the pathogenesis of IBD \[[@goz028-B60]\]. Furthermore, Mincle can recognize various lipid components released from both host cells and microorganisms released upon cell death \[[@goz028-B44], [@goz028-B61]\]. In relation to Dectin-3, we found that the development of dextran sulfate sodium (DSS)-induced colitis can be promoted by Dectin-3 deficiency in mouse models challenged with *Candida tropicalis* \[[@goz028-B46]\], supporting its role in fungal recognition in the gut. A recent study also found that Dectin-3 expression is involved in the antifungal immune response by plasmacytoid DCs to *C. neoformans* infection *in vitro* \[[@goz028-B62]\]. Another study compared the responses to pulmonary *cryptococcal* infection between Dectin-3 knockout mice and wild-type mice, and found that the murine immune responses to *C. neoformans* infection did not necessarily require Dectin-3 \[[@goz028-B63]\]. Dectin-3 can also recognize TDM and induce Mincle expression upon *Mycobacterium* infection via CARD9--Bcl-10--MALT1-dependent NF-κB activation and the increased Mincle expression can enhance the ability of host innate immune system to sense *Mycobacterium* infection \[[@goz028-B64]\]. CLRs-mediated immune response against the gut microbiota ======================================================== As discussed above, CLRs are the key PRRs in the intestinal microenvironment that mediate antifungal immunity. Both the innate and adaptive immune responses are induced and modulated by these receptors to resist infection and maintain homeostasis of the gut microbiota. The CLR-mediated innate immune responses mainly include fungal binding and phagocytosis, induction of antifungal effector mechanisms and the production of a variety of soluble mediators, such as cytokines, chemokines and inflammatory lipids \[[@goz028-B65]\]. The expression of these soluble mediators can contribute to the host adaptive immunity, which mainly involves direction and modulation of the Th1, Th2, Th17 and regulatory T cell (Treg) responses \[[@goz028-B48]\] ([Figure 2](#goz028-F2){ref-type="fig"}). {#goz028-F2} It is commonly accepted that the intestinal mucosal barrier is indispensable for protection against invasion of pathogens from the gut microbiota and for limiting their overgrowth or translocation \[[@goz028-B21]\]. The intestinal mucosa barrier not only serves as a mechanical defense barrier, but it also contains many types of mucin glycoproteins, anti-microbial peptides and secreted immunoglobulins (sIgA and IgG), all of which are involved in protecting the gastrointestinal mucosa of the host. Furthermore, M cells, which are located on the gut-associated lymphoid tissue of Peyer\'s patches and the mucosa-associated lymphoid tissue of other parts of the gastrointestinal tract lacking mucin, can serve as a portal. Pathogens, particularly *C. albicans*, can cross the barrier through this portal \[[@goz028-B66]\]. As discussed above, CLRs expressed on myeloid cells recognize pathogens in the gut and mediate the downstream immune responses. Many pro-inflammatory cytokines and chemokines, such as IL-1α/β, IL-6, G-CSF, GM-CSF, TNF-a, IL-8 and CCL20, can be produced following *C. albicans* recognition by epithelial cells \[[@goz028-B67]\]. One study found that sIgA delivery via M cells depends on Dectin-1 along with the Siglec-5 receptor, suggesting a role for Dectin-1 in the first line of defense in the gut immune response \[[@goz028-B68]\]. More recently, Dectin-3 deficiency was found to promote the development of colitis induced by severe damage to colon epithelial cells and impaired tissue repair in DSS-induced colitis models, indicating that CLRs may have additional roles in protecting the intestinal mucosa \[[@goz028-B46]\]. Intestinal epithelial cells (IECs) appear to be able to recruit lymphoid and myeloid cells to the mucosal layer \[[@goz028-B69]\], but whether this activity is related to CLRs remains unclear. The intestinal phagocytes, which are mainly located in the lamina propria, play the central role in the innate immune response against invasive pathogens once they cross the intestinal mucosal barrier \[[@goz028-B65], [@goz028-B70]\]. Phagocytes include macrophages, DCs, lymphoid cells, leukocytes and mast cells. These immune cells participate in the response to the pathogenic microorganisms and contribute to the tolerance of the normal intestinal flora. As discussed above, after being recognized by CLRs expressed on phagocytes, invasive microorganisms are internalized via the actin-dependent process of phagocytosis \[[@goz028-B71]\]. In DCs, Dectin-1 triggers phagocytosis via its cytoplasmic ITAM-like motif in a process that requires Syk \[[@goz028-B72]\]. After phagosome maturation, oxidative and non-oxidative mechanisms are used synergistically by phagocytes to kill internalized or extracellular fungi. The oxidative mechanisms include the respiratory burst and reactive nitrogen intermediates; the non-oxidative mechanism mainly depends on the production of anti-microbial peptides (AMPs), hydrolases and nutrient limitation \[[@goz028-B65]\]. Different phagocytes have various abilities to kill fungi or to limit their growth, depending on the specific fungal species. For example, neutrophils are the main cells that kill *C. albicans*, while DCs are more potent against *Histoplasma capsulatum* \[[@goz028-B73], [@goz028-B74]\]. Upon recognition of pathogens by CLRs, Syk/CARD9 induction activates NF-κB signaling, ultimately leading to the secretion of pro-inflammatory cytokines, including GM-CSF, TNF, MIP-2, IL-2, IL-10, IL-6, IL-23 and IL-1β. This signaling network initiates the differentiation of naïve T cells into Th cell lineages, including Th1, Th2, Th17 and Tregs, which are critical for antifungal immunity. Antigen presentation by the antigen-presenting cells (mainly DCs) to T cells normally occurs in the lamina propria \[[@goz028-B66]\] and Th17 differentiation is mainly promoted by induction of IL-1β, IL-6 and IL-23, which drives IL-17 and IL-22 production \[[@goz028-B75]\]. Neutrophil recruitment, epithelial AMP production and mucosal antifungal protection can also be driven by these cytokines \[[@goz028-B65]\]. However, aberrant Th17 responses may cause unexpected pathogenic hyper-inflammatory disorders that can also be tightly controlled by IL-1β production that is induced due to Dectin-1 activation \[[@goz028-B76]\]. The Th1 response, which depends on IL-12p70 and can also be triggered via a Dectin-1-directed Syk-Raf1 cooperative signaling pathway, can be shifted to a Th2 cell response by inhibition of the Raf1 or IRF1 signaling pathways \[[@goz028-B11]\]. Cooperation between the Dectin-1-induced signaling pathways is essential for the Th cell response that shapes the antifungal adaptive immune response. Unlike Dectin-1, which is capable of self-activation, Dectin-2 is activated via its association with FcRγ. Dectin-2 expressed on DCs recognizes fungal hyphae \[[@goz028-B77]\] (which consists mainly of α-mannoses) to promote NF-κB activation via Syk recruitment and subsequent activation of the CARD9--Bcl-10--MALT1 complex. The cytokines induced through this process, including IL-6, IL-23, IL-12 and IL-1β, lead to the induction of Th17 responses. A range of pro-inflammatory cytokines, including TNF, IL-6 and IL-23, as well as several anti-inflammatory cytokines, such as IL-2 and IL-10, can also be induced by Dectin-2 to promote Th17 responses \[[@goz028-B74], [@goz028-B78]\]. Depending on the DC subset and specific ligands, different signaling pathways can be activated via Mincle's association with FcRγ (similar to Dectin-2) to mediate the adaptive antifungal immune response. In human DCs, Mincle triggers the Syk-dependent CARD9--BCL-10--MALT1 complex pathway, which then results in phosphoinositide 3-kinase (PI3K)- and PKB (protein-kinase B)-dependent activation of the E3 ubiquitin ligase. The activation of the E3 ubiquitin ligase will suppress IL-12p70 production to skew Th cell differentiation in favor of Th2 cell responses \[[@goz028-B79]\]. Furthermore, Treg cells can help regulate the immune response during intestinal infection. In mouse models, the numbers of CD4^+^CD25^+^ cells and Foxp3 Tregs in the mesenteric lymph nodes (mLNs) and stomachs of infected mice are dramatically increased during intestinal *C.* *albicans* infection. CLR-mediated T cell differentiation towards Tregs also plays important roles in homeostasis in the intestinal environment \[[@goz028-B66]\]. CLRs-mediated antifungal immunity and related diseases ====================================================== The prevalence of IBD is expanding globally, now affecting over 1 million individuals in the USA and over 2.5 million individuals in Europe, as well as a significantly increased number of people in Asia, South America and the Middle East \[[@goz028-B80], [@goz028-B81]\]. IBD pathogenesis has a genetic basis and involves the immune responses activated by various environmental factors \[[@goz028-B14]\]. In the pathogenesis of IBD, immune tolerance can be reduced by the microbes, which can then result in activation of inappropriate pathways intended to protect the host against pathogen invasion \[[@goz028-B82]\]. Most studies on microbe in the gut focus on the analysis of bacteria \[[@goz028-B83]\]. Recently, the role of fungi has also attracted our attention. By using ITS2 sequencing, the fungal community in the feces of 235 patients with IBD was identified. There is a skew in the fungal microbiota of the IBD patients, with an increased *Basidiomycota/Ascomycota* ratio, a decreased proportion of *S.* *cerevisiae* and an increased proportion of *C. albicans* compared with the microbiota of healthy subjects. It is interesting that, in the gut environment of IBD, fungal growth is favored over bacteria growth \[[@goz028-B12]\]. However, after a prolonged antifungal drug treatment, mice have more severe colitis and a significant change in composition of their fungal microbiota \[[@goz028-B86]\]. Recently, studies found that the gastrointestinal microbiota and the interplay between the host immune system and the gut flora may be associated with IBD onset and exacerbation \[[@goz028-B87]\]. A polymorphic variant of Dectin-1 is related to IBD severity in patients \[[@goz028-B8]\]. Moreover, mice lacking Dectin-1 are more susceptible to colitis. *Candida* *tropicalis* contribute to a more severe colitis pathology in mice with Dectin-1 deficiency than wild-type mice. An increased production of IL-17 and IFN-γ by T cells along with increased TNF-α, IL-23p19, IL-17a and defensins detected from the colons of Dectin-1^--/--^ mice. The observation of the increased production of inflammatory mediators along with the more severe colitis in Dectin-1^--/--^ mice suggests an inability of Dectin-1 deficiency to activate effective immune responses to a specific pathogen, which promotes the inflammation and leads to more severe colitis \[[@goz028-B88]\]. This research suggests a role played by fungi to stimulate immune responses in the gut through Dectin-1 in the pathology of IBD. It is more interesting that Dectin-1 is also involved in aggravating inflammation despite its protective role against fungi pathogen. A lack of Dectin-1 signaling can facilitate the growth and expansion of *Lactobacilli*, which exerts protective effects during colitis. However, which role will Dectin-1 play depends on the situation of the gut microbiota. Overall, Dectin-1 can serve as a double-edged sword with a positive effect of protecting the host against the infection as well as a negative effect of hindering the expansion of beneficial bacteria \[[@goz028-B89], [@goz028-B90]\]. Besides Dectin-1, Dectin-3 was also involved in the pathogenesis of IBD. Mice lacking Dectin-3 are more susceptible to colitis and have higher fungal burden in the gut. Furthermore, supplemental of *C. tropicalis* promoted the pathogenesis of colitis in Dectin-3-deficiency mice \[[@goz028-B46], [@goz028-B91]\]. Cancer is associated with changes in the interactions between the gut microbiota, intestinal epithelium and host immune system \[[@goz028-B92]\]. Colorectal carcinogenesis can be affected by the gut flora through various mechanisms including specific alterations in the composition of gut microbiota and by-products that target IECs \[[@goz028-B93]\]. A study characterizing the mycobiotic signatures of 131 subjects, which were divided into polyp and colorectal-cancer (CRC) groups, along with a healthy control population for observing the fungal dysbiosis, found decreased fungal diversity. An increased *Ascomycota/Basidiomycota* ratio along with an increased growth of *Trichosporon* and *Malassezia* was also recognized. This alteration indicates that fungal dysbiosis and its related immune responses may be involved in polyp and CRC pathogenesis \[[@goz028-B94]\]. In patients with colorectal adenoma, there is also a change in the fungal microbiota structure at different stages of adenoma \[[@goz028-B95]\]. Recently, CARD9 deficiency was found to induce myeloid cell differentiation into myeloid-derived suppressor cells with increased *C. tropicalis*, which finally promotes the progression of colon cancer \[[@goz028-B96], [@goz028-B97]\]. Although the role of fungi in the pathogenesis of CRC gradually came into view, how CLRs participate in this process remains unclear to us. One study found that Dectin-1 expression is up-regulated in hepatic fibrosis and liver cancer, and Dectin-1 protects against inflammation-induced oncogenesis by suppressing TLR4 signaling \[[@goz028-B18]\]. Furthermore, Dectin-1 expressed on DCs and macrophages is critical in the recognition of and signaling by tumor cells \[[@goz028-B98]\]. In summary, CLRs are the predominant PRRs involved in recognition of the mycobiota and regulation of gastrointestinal immunity. Important roles can be played by CLRs in the development and progression of gastrointestinal illness, and much could be gained by exploring therapeutic approaches for mycobiota-related diseases. Future outlook ============== In this review, we reflected on recent studies on the roles of CLRs in the gastrointestinal environment and how CLRs interact with the immune system, as well as on diseases related to these processes. The harmonious symbiosis, which is maintained between commensal microorganisms and the host, is critical for good health. Maintenance of this relationship requires CLRs acting as PRRs to recognize potential pathogens to mediate innate and adaptive immune responses, which promote health and ameliorate disease. Future studies will provide more insight into how CLRs are involved in various diseases and how to exploit the interactions between the commensal fungal populations and immune responses in the gut to treat related diseases, including IBD and colon cancer. Furthermore, it remains unclear whether and how host immune responses, which are essential for protecting against serious life-threatening fungal infections, are related to disease progression. More work is needed to elucidate the relationship between the immune system and infectious challenges and to understand what strategies can be devised to treat the related diseases. Authors' contributions ====================== T.T.W., S.N.S. and Y.Y.H. provided the idea and designed the work; T.H.L. wrote and revised the paper; L.L. polished the English language and revised the paper. All authors reviewed and approved the final manuscript. Funding ======= This work is supported by grants from the National Natural Science Foundation of China (81572354 and 81772542 to T.W.) and the Natural Science Foundation of Jiangsu Province in China (BK20161400 to T.W.). The authors are sincerely grateful to Yu-Chen Pan, Dan Li, Jiao-Jiao Zhao and Dong-Ya Zhang for their generous instructions. Conflicts of interest ===================== None declared. | High | [
0.6666666666666661,
31.375,
15.6875
] |
The Guardian cites the state department: “We are grateful for the strongly supportive statements in response to Secretary Kerry’s speech from across the world, including Germany, France, Canada, Jordan, Egypt, Turkey, Saudi Arabia, Qatar, the United Arab Emirates and others.” Well done Theresa May for highlighting the weakness in the Kerry statement and standing proud. She is also offering a friendly olive branch to the the incoming US administration of Donald Trump. Historians should note that for eight years Obama has undermined the historical special relationship between the UK and the US. Obama knew best – pivot US foreign policy towards Asia and bring the US troops home, never mind the geopolitical consequences. The evidence is clear – Obama’s foreign policy, with ineffective foreign secretaries in Clinton and Kerry, has been a series of disasters. Is it surprising that terrorism has proliferated on Obama’s watch? The deal carries with it the seeds of its own destruction in that people smugglers have simply switched their attentions to Italy and getting people to come to Europe via Lampedusa bypassing Turkey altogether. Secondly, returned migrants simply have to give Turkish border guards small bribes of less than £50 GBP to let them go again, which means that with easy access to bogus documentation they can, like water, find new ways to re-enter the European continent. Turkey, under the deal gets £3 billion GBP a year and if it wants more all it has to do is release more migrants who can then enter Europe from Libya, via rubber boat to Lampedusa or via a Greek island. Merkel has increased the rate of housebuilding in Germany far beyond what is required for the indigenous population and is building reception hostels at a faster rate to accommodate more refugees and migrants. What conclusion is one expected to draw from all of this? This is not lost on Erdogan who can then make further incursions into Iraq and Syria and create more refugees which in turn give him leverage in his EU membership negotiations and his future discussions about money for migrants. Vladimir Putin, now facing another year of sanctions and increased American pressure via the missile shield going in to Romania plus oil prices at $37 USD a barrel also has an incentive to create more refugees and thus trouble for Europe by bearing down harder in Syria. The Austrians in the middle of this have a right to be concerned and so do the German people who were once properly led by Merkel but now need a change for the better. | Low | [
0.511434511434511,
30.75,
29.375
] |
The Department of Veterans Affairs’ in-house watchdog has demanded that the Project On Government Oversight turn over all information it has collected related to abuses and mismanagement at VA medical facilities, according to a subpoena delivered to POGO May 30. The subpoena from the VA Office of Inspector General demands all records POGO has received from current or former VA employees, as well as any other individuals, including veterans. The subpoena asks for records related to “wait times, access to care, and/or patient scheduling issues at the Phoenix, Arizona VA Healthcare System and any other VA medical facility.” POGO, which has a 33-year history of working with whistleblowers to expose government fraud, waste and abuse, wrote the IG today and refused to provide the records, most of which have come from confidential tips submitted through VAOversight.org. POGO and the Iraq and Afghanistan Veterans of America (IAVA) launched VAOversight.org on May 15 to offer potential whistleblowers a safe channel to confidentially report abuses in the VA healthcare system, which have been the focus of intense media scrutiny and congressional hearings in recent weeks. Since the website went live, about 700 people have submitted tips or aired grievances. About 25 percent of those tips have come from current or former VA staffers. POGO is reviewing the information it has received and is looking into many of the claims. In a letter sent to the IG this morning, POGO said the IG’s subpoena infringes on POGO’s constitutional “freedom of speech, freedom of press, and freedom of association rights as they relate to all whistleblowers and sources.” Some VA employees who contacted POGO and requested confidentiality said they feared retaliation if their names were divulged. Any of them could have reported their concerns to the VA inspector general. In fact, some of the employees told POGO that they had filed reports with the IG. Some people expressed a lack of confidence in the Office of the Inspector General. “The Inspector General’s demand stands opposed to POGO’s mission and to good government reform—both of which serve the public interest,” POGO Executive Director Danielle Brian said. “Our focus is squarely on investigating the problems in the VA healthcare system and trying to find some solutions. “Our mission as a public interest watchdog would be severely damaged if we violated the trust of our sources. We have faced these kinds of threats before and have never wavered. We will not violate the trust whistleblowers have placed in us by revealing their identities to anyone.” POGO’s letter to the VA IG. The IG’s subpoena. | Mid | [
0.641975308641975,
32.5,
18.125
] |
The long-term goals of the Program Project include the development of a detailed mechanistic model for actin filament branching mediated by the Arp2/3 complex and to characterize the role of the Arp2/3 complex in directed cell motility. Here, we aim at further characterizing the structures of key intermediates and their interactions along the activation pathway of the Arp2/3 complex (Aim 1). For Aim 2, we will generate high resolution, three-dimensional (3D) structures of actin networks mediated by the complex in its cellular environment. This should allow us to determine how structural details of the Arp2/3-mediated actin networks are being modified by changes in cellular dynamics. We will apply high-resolution electron cryo-microscopy (cryo-EM), electron cryotomography and image analysis to achieve these aims. | High | [
0.6967559943582511,
30.875,
13.4375
] |
Copper oxide nanoparticles promote the evolution of multicellularity in yeast. Engineered nanomaterials are rapidly becoming an essential component of modern technology. Thousands of tons of nanomaterials are manufactured, used, and subsequently released into the environment annually. While the presence of these engineered nanomaterials in the environment has profound effects on various biological systems in the short term, little work has been done to understand their consequences over long, evolutionary timescales. The evolution of multicellularity is a critical step in the origin of complex life on Earth and a unique strategy for microorganisms to alleviate adverse environmental impacts, yet the selective pressures that favor the evolution of multicellular groups remain poorly understood. Here, we show that engineered nanomaterials, specifically copper oxide nanoparticles (CuO NPs), promote the evolution of undifferentiated multicellularity in Baker's yeast (Saccharomyces cerevisiae strain Y55). Transcriptomic analysis suggests that multicellularity mitigates the negative effects of CuO NPs in yeast cells and shifts their metabolism from alcoholic fermentation towards aerobic respiration, potentially increasing resource efficiency and providing a fitness benefit during CuO NP exposure. Competition assays also confirm that the multicellular yeast possesses a fitness advantage when exposed to CuO NPs. Our results, therefore, demonstrate that nanoparticles can have profound and unexpected evolutionary consequences, underscoring the need for a more comprehensive understanding of the long-term biological impacts of nanomaterial pollution. | High | [
0.684276729559748,
34,
15.6875
] |
Searches for March Madness are stillsurvivingandadvancing, but this week brought with it a host of other stories to spur your search questions. Here’s a look at what people were searching for this week: Germanwings Flight 9525 On Tuesday, Germanwings Flight 9525 crashed in the French Alps, killing all 150 people who were onboard. In the wake of the tragedy, people around the world have turned to search to learn more about the crash and subsequent investigation. Early searches included questions about Lufthansa, Germanwings’ parent company, and about the type of plane that had crashed; search interest in the Airbus A320 family spiked 100x in the first four hours. But after investigations revealed that co-pilot Andreas Lubitz appears to have deliberately locked himself in the cockpit and flown the plane into a mountain, the questions got more specific. People asked questions like “How do you access the plane’s cockpit?” and worried: “Is it safe to fly after the Germanwings crash?” and “Is flying becoming more dangerous?” Presidential politics Election Day 2016 is more than a year-and-a-half away, but the presidential race is already underway. On Monday, Senator Ted Cruz announced his candidacy. The Texas Republican is known for his fervent opposition to the Affordable Care Act, in particular a 21-hour filibuster-style speech on the floor of the Senate in 2013 (at one point, he read aloud from Dr. Seuss’ Green Eggs and Ham). Searchers turned to the web to answer all kinds of questions about Cruz and his beliefs, sending searches for [cruz liberty], [cruz obamacare] and [cruz wiki] to spike more than a thousand percent in the last 30 days. The top questions, though, were around whether Cruz is eligible to become President because he was born in Canada. (The answer is yes, BTW.) Pop culture mania This week Zayn Malik confirmed he is leaving boy band One Direction, sending teens worldwide into a tailspin as they asked (and searched): “Why is Zayn leaving One Direction?” The British star said that he is leaving to “be a normal 22-year-old...out of the spotlight,” which may be tough given there were more than a million searches for him on Wednesday. Meanwhile, the rest of the band will go on without him—in fact, search interest in One Direction tours spiked 5x in the U.S. the day of the announcement. If freaking out about Zayn isn’t your thing, maybe freaking out about “A” is? The mysterious villain of the ABC Family show Pretty Little Liars was revealed—sorta—on Tuesday night, leading to hundreds of thousands of searches for the show. We won’t spoil it here, even though it was frankly a little hard to miss amidst the outcry. Let’s just say, the truth is out there. ;) Finally, it’s barely spring but it’s already time to start thinking about your summer festival agenda. The line-ups of both San Francisco’s Outside Lands and Chicago’s Lollapalooza music festivals were announced this week. The top festival searched in each state breaks down almost exclusively along geographic lines and leaves us wondering: Are Massachusetts residents big Elton John fans? Taking charge of your health In a New York Times op-ed, Angelina Jolie revealed that she had undergone surgery to remove her ovaries and fallopian tubes to prevent cancer. This was the second preventative surgery for Jolie (she wrote about her double mastectomy in 2013), who made this decision because she carries a mutation in her BRCA1 gene, putting her at high risk for breast and ovarian cancer. There were more than 100,000 searches for Angelina Jolie on Tuesday, and people turned to the web to ask related questions about women’s health and cancer prevention. Tip of the week Who says National Puppy Day has to be just once a year? When you’re need of a furry pick-me-up, just ask the Google app “Ok Google, show me pictures of puppies.” Smiles are practically guaranteed. | Mid | [
0.565610859728506,
31.25,
24
] |
As Always I try to make my blog cater for peeps that are just looking for concise answers, so if you scroll down there is a TL DR section below. If you are curious of my thoughts then keep reading. My knowledge of keypresses consisted of the brief summary of “you press a key and you can make something happen”. Turns out there is a bit more to it. Keypresses have three different trigger points. I’m going to put my naive understanding of them first, then what MDN says cause that is the best source of truth out there in my opinion. I also used StackOverflow. First you have keydown My understanding of a Keydown is the action of when the button is pressed, so as the key is pressed down you can set an action or trigger something from that point. Unlike the keypress event, the keydown event is fired for all keys, regardless of whether they produce a character value. from MDN To sum keydown: Keydown will trigger for all keys pressed(odd ones being ctrl, shift. delete..etc). If you press any key thats a keydown. Second you have keyup My understanding of a key up is the “release” action of the keydown action. So you can set an action upon lifting your key from a pressed state. The keydown and keyup events provide a code indicating which key is pressed, while keypress indicates which character was entered. For example, a lowercase “a” will be reported as 65 by keydown and keyup , but as 97 by keypress . An uppercase “A” is reported as 65 by all events. from MDN To sum up keyup:. It trigger upon the release of a pressed key. Don’t worry MDN can be intimidating at times. It took me a while, but they are saying a keyup will trigger with a value of what key was pressed. Also, I think MDN has already moved on tooo.. Thirdly the keypress My understanding: Keypress is the full action of just typing a key. so hit a key and let go and you have yourself a keypress buddy. The keypress event is fired when a key that produces a character value is pressed down. Examples of keys that produce a character value are alphabetic, numeric, and punctuation keys. Examples of keys that don’t produce a character value are modifier keys such as Alt , Shift , Ctrl , or Meta From MDN Ok, so my understanding of a keypress was a bit off. It appears to be including all chars that set a value and not other keys like shift, control, alt…etc that don’t produce a character. According to MDN it appears that the keypress is deprecated so you would just need to be aware of that when using keypress. I never in my life thought that someone could complicate a key press, but here we are 2020 virus is out. Try to stay inside guys ok. TL DR The onKeyDown event is triggered when the user presses a key. The onKeyUp event is triggered when the user releases a key. The onKeyPress event is triggered when the user presses & releases a key ( onKeyDown followed by onKeyUp ). from stackoverflow In my defence I’m not as technical as MDN, my theory was around more towards what the stackoverflow post said. I’ll leave you with that guys I really hope that this helps you out and helps you know that there is a bit more to a keypress than meets the eye. On a side note. I got some positive comments on my previous post. Thanks so much for the support I’ll continue posting my findings and sharing the knowledge to help myself and others. I was so chuffed in fact i made a live sample of some javaScript and keypresses right here enjoy. | Mid | [
0.5892857142857141,
33,
23
] |
Q: Most efficient way to save way points and do comparisons? I would like to know your opinion. I created an application, where users create routes and we track this route and save all the way points in the database. Then, the application does comparisons of users way points. Currently, I use a MSSQL Server, using two tables, one for Routes and the other for storing way points (with spatial data type). The comparisons are made in a stored procedure using SQL Server geographic functions such as st_distance... I have investigated other options. One that I implemented is with Oracle 11g using objects. I store all data in only one Object Table, and the way points are stored in a Varray of a type with Latitude and Longitude attributes. This way is very efficient saving and retrieving data, but gets some complicated when comparing. I'm looking for a NoSQL solution, some algorithm or method to do this efficiently. What do you think? A: Using database functions like STDistance for all n records is suboptimal. Your CPU overhead will increase exponentially. What you should do is check for the amount of points within a rectangle around the current epicenter you are searching. Here's an example (in MySQL): SELECT * FROM `points` WHERE `latitude` >= X1 AND `latitude` <= X2 AND `longitude` >= Y1 AND `longitude` <= Y2 This provides a reduced superset of points that should then be further reduced by calculating the orthodromic distance (with respect to the curvature of the Earth) using the Haversine formula. Don't forget to set up a composite index on latitude and longitude. Here it is in PHP: <?php function haversine($latitude1, $longitude1, $latitude2, $longitude2, $unit = 'Mi') { $theta = $longitude1 - $longitude2; $distance = (sin(deg2rad($latitude1)) * sin(deg2rad($latitude2))) + (cos(deg2rad($latitude1)) * cos(deg2rad($latitude2)) * cos(deg2rad($theta))); $distance = acos($distance); $distance = rad2deg($distance); $distance = $distance * 60 * 1.1515; switch ($unit) { case 'Mi': break; case 'Km': $distance = $distance * 1.609344; } return (round($distance, 2)); } ?> To recap: Here's an example image illustrating what to do: The first search would involve a bounding box collision search (MySQL example) to determine the superset, excluding the red points. The second verification process would involve calculating if the points are within an appropriate orthodromic distance with the Haversine formula (PHP example) and taking a subset (composed of the black points). | High | [
0.671490593342981,
29,
14.1875
] |
Delirium: The Nurse's Role in Prevention, Diagnosis, and Treatment. Delirium is an acute change in mental status typically caused by a medical condition. Delirium complicates care, increases safety risks, and has a negative impact on patient outcome. By identifying mental status changes early, the nurse is in a strategic position to prevent delirium in 30%-40% of at-risk patients. An interdisciplinary team approach can prevent, diagnose, and treat delirium to improve safety, reduce cost of care, and optimize patient outcomes. | High | [
0.6829268292682921,
35,
16.25
] |
Q: Print Data Frame with long string variables I have a data frame which contains descriptions of survey questions which can be quite long. I am trying to figure out a way to neatly print them. Here's an example: foo <- data.frame(v1 = 1:5, v2 = rep(c("This is a really long description of a survey question that gives a bunch of information about the question and can be very long blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah"), 5), v3 = 6:10) I would like the printed output to look something like this: v1 v2 v3 1 1 This is a really long description of 6 a survey question that gives a bunch of information about the question and can be very long blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah 2 2 This is a really long description of 7 a survey question that gives a bunch of information about the question and can be very long blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah 3 3 This is a really long description of 8 a survey question that gives a bunch of information about the question and can be very long blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah 4 4 This is a really long description of 9 a survey question that gives a bunch of information about the question and can be very long blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah 5 5 This is a really long description of 10 a survey question that gives a bunch of information about the question and can be very long blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah blah A: flextable::flextable(foo, cwidth = c(0.5,7,0.5)) | Mid | [
0.612048192771084,
31.75,
20.125
] |
Grand National Night Grand National Night is a 1953 British thriller brought to the screen by George Minter, produced by Phil C. Samuel, and based off a play written by Campbell and Dorothy Christie. It was directed by Bob McNaught and starred Nigel Patrick, Moira Lister and Beatrice Campbell (Patrick's wife) with support from Michael Hordern, Noel Purcell and a cameo role from Colin Gordon. Cinematography was by Jack Asher. Previous to this film version Grand National Night had been presented as a BBC Radio serial as well as the original stage play, which was produced in 1945 and 1946. The cast of the original play was headed by Leslie Banks as Gerald Coates Plot Racehorse trainer Gerald Coates (Nigel Patrick) kills his wife Babs (Moira Lister) during an argument. Cast Nigel Patrick as Gerald Coates Moira Lister as Babs Coates Beatrice Campbell as Joyce Penrose Betty Ann Davies as Pinkie Collins Michael Hordern as Inspector Ayling Noel Purcell as Philip Balfour Leslie Mitchell as Jack Donovan Barry MacKay as Sergeant Gibson Colin Gordon as Buns Darling Gibb McLaughlin as Morton Richard Grayden as Chandler May Hallatt as Hoskyns George Sequira as George Ernest Jay as Railway Official Russell Waters as Plainclothes Detective George Rose as Plainclothes Detective Arthur Howard as Hotel Manager Edward Evans as Garage Attendant Notes External links Category:British films Category:1950s thriller films Category:British thriller films Category:British films based on plays | Mid | [
0.5701559020044541,
32,
24.125
] |
Q: Prove $\inf\left\{1,\frac{1}{2},\frac{1}{3},\frac{1}{4},\frac{1}{5},\ldots\right\} =0$ Prove $\inf\left\{1,\frac{1}{2},\frac{1}{3},\frac{1}{4},\frac{1}{5},\ldots\right\} =0$ Definition I have to satisfy is $\inf X \leq a$, $\forall a \in X$ and if $z \leq a$, $\forall a \in X$ then $\inf X \geq z$ Let the set $X=\left\{1,\frac{1}{2},\frac{1}{3},\frac{1}{4},\frac{1}{5},\ldots\right\}$ then since the set $X \not = \emptyset$ it is bounded below by $0$ thus there exists an infimum for the set $X$ which we denote as $y$. It is clear that $y \geq 0$ now if $z \leq a$ $\forall a \in X$ then I need to show for any real number $z>0$, there is an $n$ where $\frac{1}{n}>z$ which will allow me to finish the proof. I know by archidiean property that given any $z>0$ there is an $n \in \mathbb{N}$ such that $\frac{1}{z}<n$ which doesn't give me what I want though. Any help would be appreciated to help me satisfy the second part. A: It is simply seen that $z=0$ is a lowerbound for $X$. Now suppose that $z^\prime > z$ is also a lowerbound, i.e. $\forall a \in X : z^\prime \leq a$. Now, as you suggested, there is an $n \in \mathbb{N}$ such that $0 <\frac{1}{n} < z^\prime$, which contradicts our assumption that $z^\prime$ was also a lowerbound. Hence $z$ is the largest lowerbound of $X$. | High | [
0.6682926829268291,
34.25,
17
] |
Hormone replacement therapy and cancer risk: a systematic analysis from a network of case-control studies. To provide comprehensive and quantitative information on the benefits and risks of hormone replacement therapy (HRT) on several cancer sites, we systematically examined the relation between HRT use and the risk of various cancers in women aged 45-79 by using data from a framework of case-control studies conducted in Italy between 1983 and 1999. The overall data set included the following incident, histologically confirmed neoplasms: oral cavity, pharynx, larynx and esophagus (n = 253), stomach (n = 258), colon (n = 886), rectum (n = 488), liver (n = 105), gallbladder (n = 31), pancreas (n = 122), breast (n = 4,713), endometrium (n = 704), ovary (n = 1,614), urinary bladder (n = 106), kidney (n = 102), thyroid (n = 65), Hodgkin's disease (n = 26), non-Hodgkin's lymphomas (n = 145), multiple myeloma (n = 65) and sarcomas (n = 78). The control group comprised 6,976 women aged 45-79 years, admitted for a wide spectrum of acute, nonneoplastic conditions. Odds ratios (OR) and the corresponding 95% confidence intervals (CI) for use of HRT were derived from multiple logistic regression equations. There was an inverse association between ever use of HRT and colon (OR = 0.7), rectum (OR = 0.5) and liver cancer (OR = 0.2), with a consistent pattern of protection for duration of use. An excess risk was found for gallbladder (OR = 3.2), breast (OR = 1.1), endometrial (OR = 3.0) and urinary bladder cancer (OR = 2.0). These data from a southern European population add some useful information on the risk-benefit assessment of HRT among postmenopausal women. | High | [
0.7159763313609461,
30.25,
12
] |
Microbial biomass and activity of an agricultural soil amended with the solid phase of pig slurries. Information about the mineralisation rates and effects on soil microorganisms must be obtained prior to the rational use of organic wastes in agriculture or forestry. The objective of this work was to study the mineralisation of two manures derived from the solid phase of pig slurries and the effects on the soil microbial biomass of an agricultural soil. Samples of this soil were mixed at two different rates with two manures derived from the solid phase of pig slurry (composted, CSP, and non-composted, NSP), and then were incubated during 163 days. Carbon mineralised from manures was fitted to first-order kinetic model, and small differences were found between manures despite the composting of one of them. Approximately 45% of the C added was mineralised in the experimental period. The soil microbial biomass C (C(mic)) was increased by the amendments according to the application rate. The sudden increases of the qCO(2) in the treated samples were ephemeral. The most appreciable differences between these manures were those related with net N mineralisation, being greater in the NSP-treated samples. The application of the solid phase of pig slurries, composted or not, could be a feasible practice to enhance in a short-term the microbial biomass of agricultural soils. In order to avoid an excessive release of inorganic N, the use of composted materials is preferred. | High | [
0.6783042394014961,
34,
16.125
] |
Metro Manila (CNN Philippines) — The Philippines drops 10 places in the latest World Economic Forum (WEF) Competitiveness Index. In its Global Competitiveness Report 2016-2017 released Tuesday, the WEF ranks the Philippines 57th most competitive economy out of 138. The country's overall score of 4.4 (out of 7) is the same for the past three years. The country receives its highest rank in "macroeconomic environment" at 20th (up four notches), and its lowest rank in "goods market efficiency" at 99th (down 19 notches from last year). The WEF also conducted an "Executive Opinion Survey" among the country's business leaders to identify problems in doing business. The survey was done between February and June 2016, or during the Aquino administration. The top five concerns are inefficient government bureaucracy, inadequate supply of infrastructure, corruption, tax rates, and tax regulators. Singapore is the most competitive Southeast Asian economy, maintaining its second spot in the overall index. Seven of the region's economies are in the upper-half of the index: Singapore: 2nd Malaysia: 25th Thailand: 34th Indonesia: 41st Philippines: 57th Brunei: 58th Vietnam: 60th "Declining openness in the global economy is harming competitiveness and making it harder for leaders to drive sustainable, inclusive growth," said WEF Founder and Executive Chairman Klaus Schwab. Despite the country's lower performance, the WEF groups the Philippines with Cambodia and China which posted big gains among East Asian and Pacific economies since 2007. | Mid | [
0.636363636363636,
33.25,
19
] |
Stories From Our Clinic: The Chocolate-Eating Poodle This is a true story. . . one of the interesting cases that Drs. Foster and Smith have encountered. We sincerely hope you enjoy this fascinating and educational tale. Although this story is about Christmas, we feel it applies to any holiday that involves chocolate whether it's Easter, Halloween, or even Valentine's Day! Marji, a four-year-old apricot Miniature Poodle, was a healthy and happy dog ever since we first met her as a puppy. Poodles are considered some of the most intelligent canines, and little Marji was no exception. So when the holidays came around, her owner thought nothing of stashing presents under the Christmas tree, including a small gift from her neighbor. Debbie went to work one morning and left Marji loose in the house as usual. But when she got home, Debbie sensed something strange: no greeting at the door. Worried, she went looking for her dog and discovered chewed-up wrapping paper on the floor and an open gold box. It was the gift their neighbor left - a 1-lb. box of fancy chocolates, and one-half of them were missing. Debbie found Marji standing up in her crate, looking shaky and restless. The ten-pound poodle had consumed nearly half a pound of rich, mixed chocolates, and the amount she consumed was dangerous. Chocolate contains caffeine and theobromine which are methylxanthines and can be toxic. The darker the chocolate, the more of these substances it contains. Methylxanthines in chocolate affect four areas of the pet's body: Increase the rate and force of contractions of the heart Act as a diuretic, causing the pet to lose body fluids Cause vomiting and diarrhea Cause convulsions, seizures, and even death We got the panicked call that night and advised Debbie to bring Marji in immediately. Debbie wasn't sure when the chocolates were eaten, but since Marji was already exhibiting nervous symptoms, we could deduce that the theobromines had already entered the dog's system. We also knew that we would have to get her to vomit any residual chocolate and give her activated charcoal to help neutralize what was in her intestines. We also had to control any seizures, monitor her heart, start IV fluid therapy and observe the poodle for at least 24 hours until the symptoms subsided. Luckily, we were able to treat Marji in time. She went back to her owners the next evening, and they now stow unknown gifts in a place inaccessible to Marji. Chocolate toxicity is one of the most common poisonings during the holiday season; please be careful what you put under the tree this year. What to do if your dog eats chocolate: Call your veterinarian or emergency clinic immediately. Make sure you know the approximate amount of chocolate, the type, and how long ago your pet consumed it. If the chocolate has been ingested within a short period of time, your veterinarian may have you induce your pet to vomit. The effects of chocolate toxicity may not be apparent right away, but call your veterinarian, regardless. | High | [
0.6744730679156901,
36,
17.375
] |
The photorefractive effect involves light-induced charge redistribution in a nonlinear optical material to produce internal electric fields which by virtue of the optical nonlinearity, produce local changes in the index of refraction such that dynamic, erasable holograms are formed which diffract light. The photorefractive effect is achieved by exposing the material to an optical intensity pattern consisting of bright and dark regions, such as formed by interfering two coherent laser writing beams of the same polarization. Mobile charge generated in the material migrates under the influence of diffusion and drift processes to form internal space charge electric fields which create refractive index variations due to the electrooptic effect. These variations in refractive index in the photorefractive material are known as index gratings. The index gratings diffract light and are useful for a variety of applications, including storage of holographic images, diffractive optical elements, and photorefractive two-beam coupling. Inorganic crystals exhibiting the photorefractive effect are well known in the art as described in Gunter and Huignard, "Photorefractive Materials and Their Applications", Vols. I and II ("Topics in Applied Physics", Vols. 61 and 62) (Springer, Berlin, Heidelberg, 1988). Inorganic photorefractive crystals have been fabricated into optical articles for the transmission and control (change in phase, intensity, or direction of propagation) of electromagnetic radiation, as well as for holographic image and data storage. However, it is technically difficult to fabricate such crystals into desired large area samples or thin-layered devices such as optical wave guides or multiple-layer stacks. Further, it is difficult to dope crystalline material with large concentrations of dopants in order to achieve desired property improvements, such as increase in the speed and/or magnitude of the photorefractive effect, because dopants are often excluded from the crystals during growth. Certain polymeric photorefractive materials have been described by Ducharme et al., U.S. Pat. No. 5,064,264, and Schildkraut et al., U.S. Pat. No. 4,999,809. These polymeric materials can be fabricated into thin-layered devices such as optical waveguides and multilayer stacks. Further, they can be readily doped with materials to improve a photorefractive effect. Schildkraut describes a photorefractive device having a layer of material comprising a sensitizer, a charge transporting layer, a binder, and an organic molecular dipole, which has been poled in an electric field at elevated temperatures so that the alignment of the molecular dipoles remains for long times at ambient temperatures. Although the material is shown to have light-induced changes in measured properties, Schildkraut does not show the formation of a photorefractive grating to demonstrate a photorefractive device. Ducharme et al. describe photorefractive materials comprising a polymer, a nonlinear optical chromophore, a charge transport agent, and optionally a charge generator. Although these materials are useful in certain applications, there still is a desire in the industry for a photorefractive process having longer characteristic decay time of diffraction efficiency. It is therefore an object of the present invention to provide an improved process for photorefractive index grating formation. Other objects and advantages will be apparent from the following disclosure. | High | [
0.6943396226415091,
34.5,
15.1875
] |
Clinical and functional characterization of a human ORNT1 mutation (T32R) in the hyperornithinemia-hyperammonemia-homocitrullinuria (HHH) syndrome. We studied two related families (HHH013 and HHH015) with the hyperornithinemia-hyperammonemia-homocitrullinuria (HHH) syndrome, a disorder of the urea cycle and ornithine degradation pathway, who have the same novel ornithine transporter (ORNT1) genotype (T32R) but a variable phenotype. Both HHH015 patients are doing well in school and are clinically stable; conversely, the three affected HHH013 siblings had academic difficulties and one suffered recurrent episodes of hyperammonemia and ultimately died. Overexpression studies revealed that the product of the ORNT1-T32R allele has residual function. Ornithine transport studies in HHH015 fibroblasts, however, showed basal activity similar to fibroblasts carrying nonfunctional ORNT1 alleles. We also examined two potential modifying factors, the ORNT2 gene and the mitochondrial DNA lineage (haplogroup). Haplogroups, associated with specific diseases, are hypothesized to influence mitochondrial function. Results demonstrated that both HHH015 patients are heterozygous for an ORNT2 gain of function polymorphism and belong to haplogroup A whereas the HHH013 siblings carry the wild-type ORNT2 and are haplogroup H. These observations suggest that the ORNT1 genotype cannot predict the phenotype of HHH patients. The reason for the phenotypic variability is unknown, but factors such as redundant transporters and mitochondrial lineage may contribute to the neuropathophysiology of HHH patients. | High | [
0.661577608142493,
32.5,
16.625
] |
Related Links JACKSONVILLE, Fla. -- Bacardi Bottling Corp. was cited Monday with 12 safety violations by the Occupational Safety and Health Administration after the August death of a temporary worker on his first day on the job at its Jacksonville facility. Lawrence Daquan Davis, 21, was crushed to death by a palletizer machine. He worked for Remedy Intelligent Staffing, which provides Bacardi Bottling's temporary staffing service. "A worker's first day at work shouldn't be his last day on earth," said Assistant Secretary of Labor for Occupational Safety and Health Dr. David Michaels. "Employers are responsible for ensuring the safe conditions of all their employees, including those who are temporary." Davis was cleaning glass from under the hoist of a palletizing machine when an employee restarted the machine, according to a media release from the U.S. Department of Labor. OSHA investigators found that Bacardi Bottling had not trained temporary employees on utilizing locks and tags to prevent the accidental start-up of machines and failed to ensure its own employees utilized procedures to lock or tag out machines. Two willful citations were issued because Bacardi Bottling failed to develop, document and utilize lockout/tag out procedures for the control of potentially hazardous energy and train temporary workers on lockout/tag out procedures. A willful violation is defined as one committed with intentional knowing or voluntary disregard for legal requirements or with plain indifference to worker safety and health. "Employers must train all employees, including temporary workers, on the hazards specific to that workplace -- before they start working," Michaels said, "Had Bacardi done so, this tragic loss of life could have been prevented." The company has 15 business days from receipt of the citations and proposed penalties to comply. | Mid | [
0.6327944572748261,
34.25,
19.875
] |
Q: Material-ui tooltip not working properly I'm trying to use the material-ui tooltip. I want the tooltip to be displayed at the top. Even after setting placement="top" The demo can be found here What am I doing wrong in here? A: Because page has not enough space to show tooltip on top position Just add simple padding then try it again. <div style={{padding: '50px'}}> <Tooltip placement="top" title="Chart Type"> <IconButton > <ShowChart /> </IconButton> </Tooltip> </div> | Low | [
0.48395721925133606,
22.625,
24.125
] |
Carolyn Murray I always enjoy reading Cutting Edge PR e-News, but this month’s is especially timely with what is going on in my work life. It gave me two ideas that I can use next week to improve working relations and keep moving toward our goals. Thank you so much for an informative newsletter. Anchorage, Alaska, USA About the author Kim Harrison Kim Harrison loves sharing actionable ideas and information about professional communication and business management. He has wide experience as a corporate affairs manager, consultant, author, lecturer, and CEO of a non-profit organization. Kim is a Fellow and former national board member of the Public Relations Institute of Australia, and he ran his State’s professional development program for 7 years, helping many practitioners to strengthen their communication skills. People from 115 countries benefit from the practical knowledge shared in his monthly newsletter and in the eBooks available from cuttingedgepr.com. | Mid | [
0.636165577342047,
36.5,
20.875
] |
Masks, masks everywhere… Yesterday I was surprised at the increase number of people wearing masks INSIDE of buildings. It is normal to see them wearing masks on the motorbike but the scene at the Phu Nhuan Big C was surprising. So far, only 936 people have been tested with H1N1 in Vietnam, 450 have been cured, and most importantly, no deaths. Yet there is a scare across the city. Hmm, why didn’t I think about importing masks into Vietnam when the H1N1 scare started? The funny thing is that the masks do not prevent you from contracting the virus. They prevent you from spreading it. Why do you think surgeons wear masks during surgery? Not to protect themselves from the patient but to protect the patient from infection. So I consider anyone wearing a mask to be a carrier of the virus. | Low | [
0.47692307692307606,
27.125,
29.75
] |
Q: Selenium: stale element reference: element is not attached to the page document in Python I'm using Python's version of Selenium to iterate through Select elements options. It works quite well on of the websites, but fails on the another one, with error: Message: stale element reference: element is not attached to the page document I looked it up of course, but the answers I found didn't work out for me. I use time.sleep() to wait for page to load and I can see it being loaded in the browser. I'm not sure what should I do with it. How it looks in code: options = Select(driver.find_element_by_xpath("my_element's_xpath")).options for option in options: option.click() sleep(5) First run it works fine, second run I get the error. Here is the Select element in Dev Tools in Chromium: screenshot I believe it might have to do something with first select option not having <option> tag around it, but I'm not sure how to remove it from DOM. A: The code in my program is a bit larger than what I showed, and as J0HN pointed out, it caused browser to refresh. I solved it in a kind of hack, storing each option value in references list, then iterating through it. Code speaks better than words so take a look at it below: for option in options: options_reference.append(option.text) for option in options_reference: option_element = driver.find_element_by_xpath( "//*[contains(text(), '" + option + "')]") option_element.click() It can be further improved by narrowing down XPath to option tag only. | Mid | [
0.6317460317460311,
24.875,
14.5
] |
The present invention relates to diphenylalkylamine derivatives, which have affinity for the opioid xcex4 receptor and are useful in the medicinal field, and relates to medicaments comprising said compounds as an active ingredient. Opioid receptors are mainly classified into three types, i.e., xcexc, xcex4 and xcexa from a viewpoint of differences in pharmacological actions. On the basis of the discovery of an endogenous opioid peptide in 1970""s, some progresses were made in studies about their mechanism of action. In 1990""s, studies about opioid receptor structures advanced based on genetic analysis, and their mechanism of action has been being elucidated by the molecular biology. As also for the xcex4 receptor, based on the success of cloning of xcex4 receptor by Evans, Kieffer et al. in 1992, many studies have been vigorously performed in the medicinal and pharmaceutical fields by the molecular biology. Although higher order functions of the opioid xcex4 receptors have not yet been successfully elucidated, those already reported include that an opioid xcex4 receptor agonist exhibits analgesic activity (D. E. Moulin et al., Pain, 1985, 23, 213), and that the opioid xcex4 receptor agonist has a reducing effect on adverse reactions induced by an opioid xcexc receptor agonist and an opioid xcexa receptor agonist (Gallingan et. al., J. Pharm. Exp. Ther. 1984, 229, 641). Since the opioid xcex4 receptor is known to be present widely in the central and peripheral nerve systems and considered to have a wide variety of functions, discovery of an effective and selective opioid xcex4 receptor ligands can greatly contribute to therapeutic treatments of central nerve system diseases including schizophrenia, depression, cerebral apoplexy, epilepsy, Alzheimer""s disease, and Parkinson""s disease, and peripheral nerve system diseases including pains (Exp. Opin. ther. Patents, 1999, 9, 353). Compounds related to the general formula (I) of the present invention are reported in J. Med. Chem. 1994, 37, 2125, WO93/15062, WO96/36620, WO97/10230, WO98/28270, WO98/28275 and the like. The compounds described in J. Med. Chem. 1994, 37, 2125 and WO93/15062 have very high affinity for xcex4 receptors. However, these compounds have not been used clinically, because their productions are difficult due to three asymmetric centers, which are apparent from their chemical formulas, and they have poor pharmacokinetics. Derivatives having a structure with no asymmetric center are reported in WO96/36620, WO97/10230, WO98/28270, WO98/28275 and the like. However, their affinities for the xcex4 receptor are undesirably lowered compared to the compounds described above. Thus, no compound has been reported which has a structure with no asymmetric center and high affinity for the xcex4 receptor. Further, among the compounds relevant to the compounds of the general formula (I) of the present invention, derivatives that do not have a partial structure represented by X have been reported in WO94/11337, WO97/44329, WO98/43942 and the like. However, as for these compounds, affinity for xcex4 receptor has not been reported. An aim of the present invention is to provide a substance having affinity for the opioid xcex4 receptor, in particular, to provide an effective and selective opioid xcex4 receptor ligand. A further aim is to provide a medicament useful for preventive and/or therapeutic treatment of central nerve system diseases and peripheral nerve system diseases on the basis of the features. In the specification, the term xe2x80x9copioid xcex4 receptor ligandxe2x80x9d means a compound having an ability to bind to an opioid xcex4 receptor, and comprehensively includes an agonist, antagonist, partial agonist, and inverse agonist for an opioid xcex4 receptor. In order to achieve the aim described above, the inventors of the present invention studied variety of compounds. As a result, they found that compounds represented by the following general formula (I) had high affinity for the opioid xcex4 receptor, and achieved the present invention. The present invention thus provides compounds represented by the following general formula (I): [in the formula, X represents the following group (II), (III), (IV), (V), or (VI), n represents 1, 2 or 3, R1 and R2 each independently represent a hydrogen atom, a halogen atom, a lower alkyl group which may be substituted, a lower alkenyl group which may be substituted, a lower alkoxy group which may be substituted, or a hydroxy group, or represent xe2x80x94Oxe2x80x94CH2xe2x80x94Oxe2x80x94 as xe2x80x94R1xe2x80x94R2xe2x80x94, R3 represents a hydrogen atom, a halogen atom, a lower alkyl group which may be substituted, a lower alkenyl group which may be substituted, a lower alkoxy group which may be substituted, a hydroxy group, a cyano group, an amino group which may be substituted, a carbamoyl group which may be substituted, a carboxyl group, a (substituted or unsubstituted lower alkoxy)carbonyl group, or a (substituted or unsubstituted lower alkyl)carbonyl group, R4 represents a saturated or unsaturated monocyclic or bicyclic carbocyclic group or a monocyclic or bicyclic heterocyclic group containing one or more hetero atoms selected from the group consisting of an oxygen atom, a nitrogen atom, and a sulfur atom, R5, R6, R7, R8, R9, R10, R11 and R12 each independently represent a hydrogen atom, a lower alkyl group which may be substituted, or a lower alkenyl group which may be substituted, and R3 and R4, R5 and R6, R7 and R8, and R9 and R10 may bind to each other to form a cyclic structure] or salts thereof. The present invention further provides medicaments comprising a substance selected from the group consisting of the compounds represented by the general formula (I) and pharmacologically acceptable salts thereof as an active ingredient. The preferred medicaments consist of a pharmaceutical composition comprising the substance described above and an additive for pharmaceutical preparations. These medicaments are useful for preventive treatment and/or therapeutic treatment of central nerve system diseases or peripheral nerve system diseases. The present invention further provides an opioid xcex4 receptor ligand comprising a substance selected from the group consisting of the compounds represented by the general formula (I) and pharmacologically acceptable salts thereof. The present invention still further provides use of substances selected from the group consisting of the compounds represented by the general formula (I) and pharmacologically acceptable salts thereof for manufacture of the medicaments and methods for preventive treatment and/or therapeutic treatment of central nerve system diseases or peripheral nerve system diseases, which comprises the step of administering a preventively and/or therapeutically effective amount of a substance selected from the group consisting of the compounds represented by the general formula (I) and pharmacologically acceptable salts thereof to a mammal including human. The entire disclosures of Japanese Patent Application No. 2000-085202 are incorporated by reference in the disclosures of the specification. Novel compounds of the present invention will be explained in more detail. In the specification, a xe2x80x9clower alkyl groupxe2x80x9d or a xe2x80x9clower alkoxy groupxe2x80x9d as a substituent, or a xe2x80x9clower alkyl groupxe2x80x9d or xe2x80x9clower alkoxy groupxe2x80x9d constituting a part of a substituent means an alkyl or alkoxy group in a straight or branched chain, cyclic form, or any combination thereof having 1 to 6 carbon atoms, preferably 1 to 4 carbon atoms. Examples thereof include methyl, ethyl, n-propyl, isopropyl, cyclopropyl, n-butyl, i-butyl, s-butyl, t-butyl, n-pentyl, cyclopentyl, n-hexyl, cyclohexyl, methoxy, ethoxy, n-propoxy, isopropoxy, cyclopropoxy, n-butoxy, i-butoxy, s-butoxy, t-butoxy and the like. Similarly, a xe2x80x9clower alkenyl groupxe2x80x9d as a substituent means a straight, branched, or cyclic alkenyl group having 2 to 6 carbon atoms, preferably 2 to 4 carbon atoms, and examples thereof include vinyl group, allyl group and the like. In a group containing an alkenyl moiety, the number of double bonds contained in the alkenyl moiety is not particularly limited, and a double bond contained in the alkenyl moiety may either be in Z- or E-configuration. The term xe2x80x9chalogen atomxe2x80x9d means a fluorine atom, chlorine atom, bromine atom, or iodine atom unless otherwise specifically mentioned. The term xe2x80x9chetero atomxe2x80x9d means, for example, a hetero atom such as an oxygen atom, nitrogen atom, or sulfur atom unless otherwise specifically mentioned, preferably an oxygen atom, nitrogen atom, or sulfur atom. A xe2x80x9cheterocyclic ringxe2x80x9d may contain two or more hetero atoms as ring-constituting atoms. In such compounds, two or more hetero atoms may be the same or different. A heterocyclic group means a residue of a heterocyclic ring obtained by removing one or more hydrogen atoms that bind to ring-constituting atoms. In the formula (I), R5 and R6 in the group (II), R7 and R8 in the group (III), and R9 and R10 in the group (TV), which groups are represented by X, may each independently bind to each other to form a cyclic structure. Examples of the ring include aziridine, azetidine, pyrrolidine, or piperidine, and an example of a preferred ring includes piperidine. An unsaturated bond may exist in a part of these rings. Further, R5, R6, R7, R8, R9, R10, R11, and R12 in the group (II), (III), (IV), (V), or (VI) represented by X preferably each independently represent a hydrogen atom or a lower alkyl group, or R5 and R6 preferably bind to each other to form piperidine. The group represented by X is preferably the group (II). The integer represented by n is preferably 2 or 3. R1 and R2 preferably represent a hydrogen atom, a halogen atom, a lower alkoxy group, or a hydroxy group, or represent xe2x80x94Oxe2x80x94CH2xe2x80x94Oxe2x80x94 as xe2x80x94R1xe2x80x94R2xe2x80x94. R1 and R2 most preferably represent a hydrogen atom, a fluorine atom, a methoxy group, or a hydroxy group. R3 is preferably a hydrogen atom or a lower alkylcarbonyl group, and most preferably a hydrogen atom. Examples of the carbocyclic ring constituting the saturated or unsaturated monocyclic or bicyclic carbocyclic group represented by R4 include rings of cyclopentane, cyclohexane, benzene, indane, naphthalene and the like, and a preferred example includes benzene. Examples of the heterocyclic ring constituting the monocyclic or bicyclic heterocyclic group containing one or more hetero atoms represented by R4 include rings of imidazole, benzofuran, indole, benzothiophene, benzothiazole, benzoxazole, benzimidazole, benzotriazole, benzisothiazole, benzisoxazole, quinoline, isoquinoline, quinazoline, pyridinoimidazole, benzoxazine and the like, and preferred examples include indole and benzimidazole. An unsaturated bond as a part of the saturated or unsaturated monocyclic or bicyclic carbocyclic group or the monocyclic or bicyclic heterocyclic ring containing one or more hetero atoms represented by R4 may be hydrogenated to form a saturated bond, or may be oxidized to form a cyclic ketone, cyclic amide (lactam), cyclic ester (lactone) or cyclic ureide structure. A substituting position of the adjacent piperidine ring may be any substitutable position. One or more hydrogen atoms on the saturated or unsaturated monocyclic or bicyclic carbocyclic group or the monocyclic or bicyclic heterocyclic ring containing one or more hetero atoms represented by R4 may be substituted. Examples of the substituent include a lower alkyl group such as methyl group, a lower alkoxy group such as methoxy group, a lower alkenyl group such as allyl group, a halogen atom, a hydroxy group, a cyano group, an amino group, a N-(lower alkyl)amino group such as N-methylamino group, a N,N-di(lower alkyl)amino group such as N,N-dimethylamino group, a nitro group, a carbamoyl group, a N-(lower alkyl)carbamoyl group such as N-methylcarbamoyl group, a N,N-di(lower alkyl)carbamoyl group such as N,N-dimethylcarbamoyl group, a carboxyl group, a lower alkoxycarbonyl group such as methoxycarbonyl group, a lower alkylcarbonyl group such as acetyl group, an oxo group, a benzyl group, a hydroxymethyl group and the like. Preferred substituents are a lower alkyl group, an oxo group, a benzyl group, and a hydroxymethyl group, and particularly preferred substituents are a lower alkyl group, an oxo group, and hydroxymethyl group. When two or more substituents exist, they may be the same or different. Positions of substituents are not limited, and they can exist at any substitutable positions. Further, an example of the compounds wherein R5 and R6 bind to each other to form a cyclic structure include the compounds in which a spiro ring is formed. Specifically, examples include the following groups (VII) and (VII) (In the following chemical formulas, a spiro ring is formed in each upper ring. In the formulas, each of two solid lines drawn from a ring represents a single bond that binds to the 3- or 5-position of the piperidine ring on which R3 and R4 substitute). [in the groups, R13 and R15 each independently represent a hydrogen atom, a lower alkyl group which may be substituted, or a lower alkenyl group which may be substituted, and R14 and R16 each independently represent a hydrogen atom, a halogen atom, a substituted or unsubstituted lower alkyl group, a lower alkenyl group which may be substituted, a lower alkoxy group which may be substituted, a hydroxy group, a cyano group, an amino group which may be substituted, a nitro group, a carbamoyl group which may be substituted, a carboxyl group, a (substituted or unsubstituted lower alkoxy)carbonyl group, or (substituted or a unsubstituted lower alkyl)carbonyl group] The definition that a group represented by R1, R2, R3, R4, R5, R6, R7, R8, R9, R10, R11, R12, R13, R14, R15 or R16 xe2x80x9cmay be substitutedxe2x80x9d means that the group may have any kind of one or more of substituents. When the groups have two or more substituents, they may be the same or different. The positions of the substituents are not limited, and they can exist at any substitutable positions. Kinds of the substituents are not limited, and examples include a lower alkyl group such as methyl group, a lower alkoxy group such as methoxy group, a lower alkenyl group such as allyl group, a halogen atom, a hydroxy group, a cyano group, an amino group, a N-(lower alkyl)amino group such as N-methylamino group, a N,N-di(lower alkyl)amino group such as N,N-dimethylamino group, a nitro group, a carbamoyl group, a N-(lower alkyl)carbamoyl group such as N-methylcarbamoyl group, a N,N-di(lower alkyl)carbamoyl group such as N,N-dimethylcarbamoyl group, a carboxyl group, a (lower alkoxy)carbonyl group such as methoxycarbonyl group, a (lower alkyl)carbonyl group such as acetyl group, and a saturated or unsaturated 3- to 6-membered carbocyclic group such as cyclopropyl, cyclopentyl, cyclohexyl and phenyl (these carbocyclic groups may have one or more substituents, and examples of the substituents include a lower alkyl group such as methyl group, a lower alkoxy group such as methoxy group, a lower alkenyl group such as allyl group, a halogen atom, a hydroxy group, a cyano group, an amino group, a N-(lower alkyl)amino group such as N-methylamino group, a N,N-di(lower alkyl)amino group such as N,N-dimethylamino group, a nitro group, a carbamoyl group, a N-(lower alkyl)carbamoyl group such as N-methylcarbamoyl group, a N,N-di(lower alkyl)carbamoyl group such as N,N-dimethylcarbamoyl group, a carboxyl group, a (lower alkoxy)carbonyl group such as methoxycarbonyl group, a (lower alkyl)carbonyl group such as acetyl group and the like), and a phenyl group is preferred. Among the compounds represented by the general formula (I), examples of preferred class of compounds include those wherein R4 represents a residue of a ring selected from a group consisting of benzene, indole, and benzimidazole (a part of unsaturated bonds of the ring may be hydrogenated to form a saturated bond(s), or the ring may be substituted). Another preferred class of compounds include those wherein R3 and R4 form a cyclic structure and represent the following group (VII) or (VIII): [in the groups, R13, R15, R14 and R16 have the same meanings as defined above]. More preferred class of compounds include those wherein X represents the group (II), (III), (IV), (V) or (VI), n is 2 or 3, R1 and R2 each independently represent a hydrogen atom, a halogen atom, a lower alkoxy group, or a hydroxy group, or R1 and R2 represent xe2x80x94Oxe2x80x94CH2xe2x80x94Oxe2x80x94 as xe2x80x94R1xe2x80x94R2xe2x80x94, R3 is a hydrogen atom or a lower alkylcarbonyl group, R4 is a residue of a ring selected from the group consisting of benzene, indole, and benzimidazole (a part of unsaturated bonds of the ring may be hydrogenated to form a saturated bond(s), and a hydrogen atom on the ring may be substituted with an oxo group, a lower alkyl group, a hydroxymethyl group, or a benzyl group), or R3 and R4 bind to each other to form a cyclic structure of a ring selected from imidazole, N-phenylimidazolidine, and isoquinoline (a hydrogen atom on the ring may be substituted with an oxo group or a lower alkyl group), and R5, R6, R7, R8, R9, R10, R11, and R12 each independently represent a hydrogen atom or a lower alkyl group, or R5 and R6 bind to each other to form piperidine. Further preferred class of compounds include those wherein X represents the group (II). In the present invention, among the compounds represented by the general formula (I), particularly preferred compounds are as follows. 1. 1-[3-(4-Diethylcarbamoylphenyl)-3-(3 -methoxyphenyl)propyl]-4-(1,3-dihydro-2H-benzimidazol-2 -on-1-yl)piperidine 2. 1-[3-(4-Diethylcarbamoylphenyl)-3-(3 -hydroxyphenyl)propyl]-4-(1,3-dihydro-2H-benzimidazol-2 -on-1-yl)piperidine hydrochloride 3. 8-[3-(4-Diethylcarbamoylphenyl)-3-(3-methoxyphenyl)propyl]-1-phenyl-1,3,8-triazaspiro[4,5]decan-4-one 4. 1-[3-(4-Diethylcarbamoylphenyl)-3-(3-methoxyphenyl)propyl]-4-(2-hydroxymethyl-1H-benzimidazol-1-yl)piperidine 5. 1xe2x80x2-[3-(4-Diethylcarbamoylphenyl)-3-(3-methoxyphenyl)propyl]-2,3-dihydro-5-methylspiro[isoquinoline-4(1H),4xe2x80x2-piperidin]-1-one 6. 1xe2x80x2-[3-(4-Diethylcarbamoylphenyl)-3-(3-hydroxyphenyl)propyl]-2,3-dihydro-5-methylspiro[isoquinoline-4(1H),4xe2x80x2-piperidin]-1-one hydrochloride 7. 1-[3-(4-Diethylcarbamoylphenyl)-3-(2-methoxyphenyl)propyl]-4-(1,3-dihydro-2H-benzimidazol-2-on-1-yl)piperidine 8. 8-[3-(4-Diethylcarbamoylphenyl)-3-(2-methoxyphenyl)propyl]-1-phenyl-1,3,8-triazaspiro[4,5]decan-4-one 9. 1xe2x80x2-[3-(4-Diethylcarbamoylphenyl)-3-(2-methoxyphenyl)propyl]-2,3-dihydro-5-methylspiro[isoquinoline-4(1H),4xe2x80x2-piperidin]-1-one 10. 1-[3-(4-Diethylcarbamoylphenyl)-3-(2-methoxyphenyl)propyl]-4-(2-hydroxymethyl-1 H-benzimidazol-1-yl)piperidine 11. 1-[3-(4-Diethylcarbamoylphenyl)-3-(3-methoxyphenyl)propyl]-4-(1H-benzimidazol-1-yl)piperidine 12. 1-[3-(4-Diethylcarbamoylphenyl)-3-(4-methoxyphenyl)propyl]-4-(1,3-dihydro-2H-benzimidazol-2-on-1-yl)piperidine 13. 1-[3-(4-Diethylcarbamoylphenyl)-3-(4-hydroxyphenyl)propyl]-4-(1,3-dihydro-2H-benzimidazol-2-on-1-yl)piperidine hydrochloride 14. 1-[3-(4-Diethylcarbamoylphenyl)-3-(3-methoxyphenyl)propyl]-4-(2-methyl-1H-benzimidazol-1-yl)piperidine 15. 1-[3-(4-Diethylcarbamoylphenyl)-3-(2-hydroxyphenyl)propyl]-4-(1,3-dihydro-2H-benzimidazol-2-on-1-yl)piperidine hydrochloride 16. 1-[3-(3-Diethylcarbamoylphenyl)-3-(3-methoxyphenyl)propyl]-4-(1,3-dihydro-2H-benzimidazol-2-on-1-yl)piperidine 17. 1-[3-(4-Diethylcarbamoylphenyl)-3-(3-methoxyphenyl)propyl]-4-(3-benzyl-1,3-dihydro-2H-benzimidazol-2-on-1-yl)piperidine 18. 1-[3-(4-Diethylcarbamoylphenyl)-3-(3-methoxyphenyl)propyl]-4-(3-cyclopropylmethyl-1,3-dihydro-2 H-benzimidazol-2-on-1-yl)piperidine 19. 1-[4-(4-Diethylcarbamoylphenyl)-4-(3-methoxyphenyl)butyl]-4-(1,3-dihydro-2H-benzimidazol-2-on-1-yl)piperidine 20. 1-[3-(4-Diethylcarbamoylphenyl)-3-(3-methoxyphenyl)propyl]-4-(3,3-dimethyl-2,3-dihydro-1 H-indol-2-on-1-yl)piperidine 21. 1-[3-(4-Diethylcarbamoylphenyl)-3-(3-fluorophenyl)propyl]-4-(1,3-dihydro-2H-benzimidazol-2-on-1-yl)piperidine 22. 8-[3-(4-Diethylcarbamoylphenyl)-3-(3-hydroxyphenyl)propyl]-8-phenyl-1,3,8-triazaspiro[4,5]decan-4-one hydrochloride 23. 1-[3-(4-Diethylcarbamoylphenyl)-3-(1,3-benzodioxole-5-yl)propyl]-4-(1,3-dihydro-2H-benzimidazol-2-on-1-yl)piperidine 24. 1-[4-(4-Diethylcarbamoylphenyl)-4-(3-hydroxyphenyl)butyl]-4-(1,3-dihydro-2H-benzimidazol-2-on-1-yl)piperidine hydrochloride 25. 1-[3-(4-Diethylcarbamoylphenyl)-3-(3,4-dihydroxyphenyl)propyl]-4-(1,3-dihydro-2H-benzimidazol-2-on- 1-yl)piperidine hydrochloride 26. 1-[3-(4-Diethylcarbamoylphenyl)-3-(3-methoxyphenyl)propyl]-4-1H-indol-3-yl)-piperidine 27. 1-[3-(4-Diethylcarbamoylphenyl)-3-(3-hydroxyphenyl)propyl]-4-(1H-indol-3-yl)-piperidine hydrochloride 28. 1-[3-(4-Diethylcarbamoylphenyl)-3-(3-hydroxyphenyl)propyl]-4-(2-hydroxymethyl-1H-benzimidazol-1-yl)piperidine hydrochloride 29. 1-[3-(4-Diethylcarbamoylphenyl)-3-phenylpropyl]-4-(1,3-dihydro-2H-benzimidazol-2-on-1-yl)piperidine 30. 1-[3-(4-Diisopropylcarbamoylphenyl)-3-(3-methoxyphenyl)propyl]-4-(1,3-dihydro-2H-benzimidazol-2-on-1-yl)piperidine 31. 1-[3-(4-Diethylcarbamoylphenyl)-3-(4-fluorophenyl)propyl]-4-(1,3-dihydro-2H-benzimidazol-2 -on-1-yl)piperidine 32. 1-[3-(4-Piperidinocarbonylphenyl)-3-(3-methoxyphenyl)propyl]-4-(1,3-dihydro-2H-benzimidazol-2-on- 1-yl)piperidine 33. 1-[3-(4-Diethylcarbamoylphenyl)-3-(3-hydroxyphenyl)propyl]-4-(1H-benzimidazol-1-yl)piperidine hydrochloride 34. 1-[3-(4-Carboxylphenyl)-3-(3-methoxyphenyl)propyl]-4-(1,3-dihydro-2H-benzimidazol-2-on-1-yl)piperidine hydrochloride 35. 4-(1,3-Dihydro-2H-benzimidazol-2-on-1-yl)-1-[3-(4-methoxycarboxylphenyl)-3-(3-methoxyphenyl)propyl]piperidine 36. 4-Acetyl-1-[3-(4-diethylcarbamoylphenyl)-3-(3-methoxyphenyl)propyl]-4-phenylpiperidine 37. 1xe2x80x2-[3-(4-Diethylaminosulfonylphenyl)-3-(3-methoxyphenyl)propyl]-2,3-dihydro-5-methylspiro[isoquinoline-4(1H),4xe2x80x2-piperidin]-1-one 38. 2,3-Dihydro-5-methyl-1xe2x80x2-[3-[4-(1-methylbutyryl)phenyl]-3-(3-methoxyphenyl)-propyl]spiro[isoquinoline-4(1H),4xe2x80x2-piperidin]-1-one 39. 1-[3-(4-Diethylaminomethylphenyl)-3-(3-methoxyphenyl)propyl]-4-(1H-benzimidazol-1-yl)piperidine Methods for preparing the novel compounds of the present invention will be explained in more detail. The novel compounds of the present invention can be prepared by the methods described below. [in the formulas, R1, R2, R3, R4, X and n have the same meanings as defined in the general formula (I), and W represents a halogen atom excluding a fluorine atom or represents a leaving group such as p-toluenesulfonyloxy group, methanesulfonyloxy group, or trifluoromethanesulfonyloxy group]. The compounds (IX) can be prepared by the method described in WO97/10230 with partial modification, and a specific preparation thereof is described later in Reference Example 1. The compounds (X) can be obtained as commercially available reagents, or can also be obtained by a known method or an improved method thereof. The compounds (I) according to the present invention can be obtained by a reaction of a compound (IX) and a compound (X) in a solvent that is not involved in the reaction (for example, dichloromethane, tetrahydrofuran, methyl ethyl ketone, N,N-dimethylformamide, dimethyl sulfoxide and the like) in the presence of a base (for example, pyridine, triethylamine, diisopropylethylamine, 4-dimethylaminopyridine, potassium carbonate, sodium carbonate and the like) at a reaction temperature of which lower limit is 20xc2x0 C. and upper limit is 100xc2x0 C., preferably lower limit is 20xc2x0 C. and upper limit is 50xc2x0 C., for a reaction time of which lower limit is 2 hours and upper limit is 48 hours, preferably lower limit is 16 hours and upper limit is 24 hours. In the synthesis of the compounds of the present invention, purification of a target compound from a reaction mixture is performed by methods ordinarily used in the filed of organic chemistry, for example, by distribution of reaction products between water and an organic solvent that is not freely miscible with water (e.g., benzene, toluene, ethyl acetate, butyl acetate, methyl isobutyl ketone, chloroform, dichloromethane and the like) for extraction, and then by concentration, crystallization and the like. Further, as required, fractionation purification by column chromatography using alumina, silica gel or the like, for example, may also be performed. The compounds of the present invention may be in the form of a salt. The salt may be an acid addition salt such as salts with inorganic acids including hydrochloric acid, nitric acid, hydrobromic acid, and sulfuric acid, salts with aliphatic monocarboxylic acids, dicarboxylic acids, hydroxyalkanoic acids, hydroxydialkanoic acids, amino acids and the like, or salts deriving from non-toxic organic acids such as aromatic acids, aliphatic acids, and aromatic sulfonic acid. Examples of such acid addition salts include hydrochloride, hydrobromide, nitrate, sulfate, hydrogensulfate, hydrogenphosphate, dihydrogenphosphate, acetate, propionate, tartrate, oxalate, malonate, succinate, fumarate, maleate, mandelate, benzoate, phthalate, methanesulfonate, benzenesulfonate, toluenesulfonate, citrate, lactate, malate, glycolate, trifluoroacetate and the like. Typical compounds of the present invention are specifically explained in detail in the examples of the present specification. Therefore, those skilled in the art can prepare any compound falling within the scope of the general formula (I) based on explanations of the general preparation methods describer above and examples described later by appropriately choosing starting compounds, reagents, reaction conditions and the like, and if necessary, applying appropriate modifications or alterations to the methods disclosed in the examples. As well as the compounds in free form or salts thereof, any hydrates and solvates thereof also fall within the scope of the present invention. The types of solvents that form the solvates are not particularly limited. Examples include solvents such as methanol, ethanol, acetone, and diethyl ether. However, the solvents are not limited to these examples. The compounds of the present invention may have one or more asymmetric carbon atoms depending on the type of substituent, and any of stereoisomers such as optically active isomers or diastereoisomers in a pure form, any mixtures of the stereoisomers, racemates and the like also fall within the scope of the present invention. The compounds of the present invention are characterized to have affinity for opioid xcex4 receptor. Therefore, the compounds of the present invention are useful for preventive and/or therapeutic treatment of central nerve system diseases such as schizophrenia, depression, cerebral apoplexy, epilepsy, Alzheimer""s disease, and Parkinson""s disease and peripheral nerve system diseases such as pains, in which the opioid xcex4 receptor is involved. The medicaments provided by the present invention are characterized to comprise at least one kind of the compound represented by the general formula (I) or pharmacologically acceptable salt thereof as an active ingredient. The medicaments of the present invention can be administered to human or animals other than human by any of oral or parenteral routes (for example, intravenous injection, intramuscular injection, subcutaneous administration, rectal administration, percutaneous administration, intraspinal administration). As the medicaments of the present invention, the substances as active ingredients, per se, may be administered. It is generally preferable to prepare and administer a pharmaceutical composition, as a form suitable for the administration route, by using one or more kinds of additives for pharmaceutical preparations. Specifically, examples of orally available formulations include tablets, capsules, powders, granules, syrups and the like. Examples of parenteral formulations include injections such as intravenous and intramuscular injections, formulations for rectal administration, oily suppositories, aqueous suppositories and the like. These various pharmaceutical preparations can be prepared by using additives for pharmaceutical preparations which are ordinarily used, for example, excipients, disintegrating agents, binders, lubricants and coloring agents. Examples of the excipients include lactose, glucose, cornstarch, sorbit, crystalline cellulose and the like. Examples of the disintegrating agents include starch, sodium alginate, gelatin powder, calcium carbonate, calcium citrate, dextrin and the like. Example of the binders include dimethylcellulose, polyvinyl alcohol, polyvinyl ether, methylcellulose, ethylcellulose, gum arabic, gelatin, hydroxypropylcellulose, polyvinylpyrrolidone and the like. Examples of the lubricants include talc, magnesium stearate, polyethylene glycol, hydrogenated vegetable oil and the like. Further, the pharmaceutical preparations can be prepared with addition of a buffer, pH modifier, stabilizer or the like as required. Although content of the compound of the present invention in the pharmaceutical composition may vary depending on types of formulations. Generally, its lower limit is about 0.1% by weight and upper limit is 50% by weigh, preferably lower limit is 0.5% by weight and upper limit is 20% by weight based on the total composition. A dose may appropriately be determined depending on each case in consideration of the age, body weight, sex, type of a disease, severity of symptoms of a patient and the like. Generally, its lower limit is 1 mg and upper limit is 1000 mg, preferably its lower limit is 1 mg and upper limit is 300 mg, per day for an adult. The dose is administered once a day or several times a day dividedly. | Low | [
0.523076923076923,
29.75,
27.125
] |
--- abstract: | A Conditional Tree Pattern ($\CTP$) expands an XML tree pattern with labels attached to the descendant edges. These labels can be XML element names or Boolean $\CXP$’s. The meaning of a descendant edge labelled by $A$ and ending in a node labelled by $B$ is a path of child steps ending in a $B$ node such that all intermediate nodes are $A$ nodes. In effect this expresses the *until B, A holds* construction from temporal logic. This paper studies the containment problem for $\CTP$. For tree patterns ($\TP$), this problem is known to be $\coNP$-complete. We show that it is $\PSPACE$-complete for $\CTP$. This increase in complexity is due to the fact that $\CTP$ is expressive enough to encode an unrestricted form of label negation: ${*}\setminus a$, meaning “any node except an $a$-node”. Containment of $\TP$ expanded with this type of negation is already $\PSPACE$-hard. $\CTP$ is a positive, forward, first order fragment of Regular XPath. Unlike $\TP$, $\CTP$ expanded with disjunction is not equivalent to unions of $\CXP$’s. Like $\TP$, $\CTP$ is a natural fragment to consider: $\CTP$ is closed under intersections and $\CTP$ with disjunction is equally expressive as positive existential first order logic expanded with the until operator. address: - '[A]{}IDSIA - Dalle Molle Institute for Artificial Intelligence' - '[b]{}FireEye' - 'ISLA, University of Amsterdam' - author: - Alessandro Facchinia - Yoichi Hiraib - Maarten Marxc - Evgeny Sherkhonovd bibliography: - 'main-bib.bib' title: Containment for Conditional Tree Patterns --- [^1] Introduction ============ Tree Patterns, abbreviated as $\TP$, are one of the most studied languages for XML documents and used in almost all aspects of XML data managment. Tree patterns are a natural language: over trees, unions of tree patterns are equally expressive as positive first order logic [@bene:stru05]. Also, like relational conjunctive queries, the semantics of $\TP$’s can be given by embeddings from patterns to tree models [@amer:tree02]. Equivalence and containment of $\TP$’s is decidable in $\PTIME$ for several fragments [@amer:tree02], and $\coNP$ complete in general [@mikl:cont04]. Tree patterns can be represented as trees, as in Figure \[fig:tp\], or be given a natural XPath-like syntax. ![The tree pattern corresponding to the XPath expression `/p[t]//s[.//v].` The node in the square box denotes the output node.[]{data-label="fig:tp"}](tpbox) ![Conditional tree pattern corresponding to . []{data-label="fig:CTP1"}](condpattern){width="30.00000%"} In this paper, we study the expansion of tree patterns with the conditional descendant axis. We call this expansion *Conditional Tree Patterns*, abbreviated as $\CTP$. Where the descendant axis in $\TP$ can be written as the transitive reflexive closure of the XPath step $\mathtt{child{::}*}$, the conditional descendant axis is the transitive closure of $\mathtt{child::p[F_1][F_2]\ldots[F_n]}$, where $n\geq 0$ and each $F_i$ is an XPath expression which might contain the conditional descendent axis itself, followed by a child step. Syntactically, the expansion is straightforward: in the tree representation add labels representing conditional tree patterns themselves to the edges. Figure \[fig:CTP1\] contains an example which is equivalent to the following XPath-like formula in which we use $(\cdot)^*$ to denote the reflexive transitive closure of a path formula: $$\begin{aligned} \mathtt{se}&\mathtt{lf::p[child::t]/(child::r)^*/child::s} \nonumber\\ & \mathtt{\ \ [ (child::a[(child::a)^*/child::b])^*/child::v]} \label{eq:CTP1}\end{aligned}$$ The edge in Figure \[fig:CTP1\] labeled by $r$ corresponds to the expression $\mathtt{(child::r)^*/child::*}$, and merely states that all intermediate nodes are labeled by $r$. The other labeled edge shows a nesting of patterns, corresponding to a nested transitive closure statement: all intermediate nodes have to be labeled by $a$ and moreover have to start an $a$-labeled path ending in a $b$-node. Conditional tree patterns are the forward fragment of Conditional XPath [@marx:cond05] without disjunction and negation. The conditional descendant axis is closely related to the strict until operator from temporal logic [@marx:cond05; @libk:reas10]. Main results {#main-results .unnumbered} ------------ Our main results concern the expressivity of $\CTP$ and the complexity of the containment problem. We consider two types of models: the standard XML trees in which each node has exactly one label, and trees in which nodes can have an arbitrary number of labels. These latter, called *multi-labeled trees*, are the models considered in temporal logic. All our results hold for both semantics. Models with multiple labels are a convenient logical abstraction for reasoning about tree patterns expanded with attribute value equalities. These are expressions of the form $@_a=c$, where $a$ is an attribute, $c$ a constant, meaning that it holds at a node if and only if the value of the $a$-attribute of the node equals $c$. (With that we can express XPath formulas like `//table[@border = ’1’]`). Containment of tree patterns has been studied in [@amer:tree02; @mikl:cont04; @neve:comp06]. The most relevant results for this paper are that containment of $\TP$’s is $\coNP$-complete in general [@mikl:cont04] and $\PSPACE$-complete when the domain of labels is finite [@neve:comp06]. We show that containment of $\CTP$’s is $\PSPACE$-complete (with both finite and infinite domain of labels). Interestingly, this increase in complexity is due to the fact that $\CTP$ is expressive enough to encode an unrestricted form of label negation: ${*}\setminus a$, meaning “any node except an $a$-node". We show that containment of $\TP$’s expanded with this form of negation is already $\PSPACE$-hard. The matching upper bound for $\CTP$ containment is easily obtained by a translation into Existential $\CTL$ [@kupf:auto00]. As a contrast we consider the expansion of $\TP$ with a *safe* form of propositional negation $n\setminus a$ , which selects nodes with the label containing $n$ and not $a$, instead of ${*}\setminus a$ [@bara:guar11]. Note that this construct only makes sense on models with multiple labels. With respect to expressivity, we show that most results for $\TP$ can be generalized to $\CTP$. $\CTP$’s can be interpreted in trees by generalizing the $\TP$-embeddings to simulations known from temporal logic. Similarly to the characterization for $\TP$ in [@bene:stru05], we show that $\CTP$’s with disjunction and union are equally expressive as positive first order logic expanded with an *until* operator. From this we obtain that like $\TP$’s, $\CTP$’s are closed under taking intersections. Organization {#organization .unnumbered} ------------ The paper is organized as follows. This section is continued with a few more comparisons between $\TP$ and $\CTP$, related work and a motivating example. Section \[sec:prelim\] contains preliminaries. Section \[sec:expr\] contains the expressivity results and Section \[sec:cont\] the results on the complexity of the containment problem. We end with conclusions and open questions. Comparing logical properties of $\TP$ and $\CTP$ {#comparing-logical-properties-of-tp-and-ctp .unnumbered} ------------------------------------------------ A characteristic difference between $\TP$ and Relational Conjunctive Queries ($\CQ$) is the disjunction property: if A $\models B\vee C$, then $A \models B$ or $A\models C$. This holds for $\CQ$, but not for $\TP$. A counterexample is $//p\models /p\ \vee\ /{*}//p$. The languages $\TP$ and $\CTP$ differ on the following: Unions : $\TP$ expanded with disjunction is equally expressive as unions of $\TP$ [@bene:stru05]. However, $\CTP$ with disjunction is more expressive than unions of $\CTP$. Countermodels : If containment between two $\TP$’s fails, there is a countermodel for it of polynomial size [@mikl:cont04]. Countermodels for $\CXP$ containment may be exponential. Complexity : $\TP$ containment is $\coNP$-complete [@mikl:cont04] and $\CTP$ containment is $\PSPACE$-complete. For both languages this remains true if we add disjunction to the language (for $\TP$ see [@neve:comp06]). However, there are a few useful technical results that $\TP$ and $\CTP$ share. Monotonicity : $\TP$ and $\CTP$ formulas are preserved under extensions of models at the leaves (i.e. when the original model is a sub*tree* of the extension). This means that if a $\TP$ ($\CTP$) formula holds in a tree, then it holds in the extensions of the tree. Multiple output nodes : The containment problem for $\TP$ and $\CTP$ formulas with multiple output nodes can be reduced to containment of Boolean $\TP$ and $\CTP$ formulas (i.e. when the formula has a single output node which is the root). Containment for unions : Containment for unions of $\TP$s can be reduced in $\PTIME$ to checking a set of containments between $\TP$ formulas [@mikl:cont04]. This reduction can be seen as a weak disjunction property. A similar property holds for $\CTP$ (see Proposition \[prop:mik\]). Multi-labeled models : There is a $\PTIME$ reduction from the containment problem over trees in which each node has exactly one label to the containment problem over multi-labeled trees. This holds for both $\TP$ and $\CTP$. For $\TP$, most of these results are in [@mikl:cont04]. Here we show how their proofs can be generalized to $\CTP$. Motivation {#motivation .unnumbered} ---------- Tree patterns exhibit a nice tradeoff between expressive power and the complexity of static analysis. However, there are natural scenarios where tree patterns are not powerful enough, e.g. in Example \[ex:cond\]. The conditional axis gives us more querying capabilities while preserving some of the nice properties of tree patterns (see the comparison above). \[ex:cond\] Conditional tree patterns are used to query tree shaped structures. As an example, take the tree structure of the UNIX file system. In this file system, every file and directory has different access permissions (read, write or execute) for different type of users (the user, the group and others). Thus, the file system can be modeled as a tree where each node corresponds to a directory or a file (labeled by “dir” and “file” respectively) and has a required attribute for each pair $(user, access\ right)$ which takes values from $\{0,1\}$. A file can only be a leaf in the tree. Assume we want to ask for all the files that are readable by the user. This means that we are looking for precisely those files for which the following permissions hold: - the file is readable for the user, - the directory in which the file resides is both readable and executable for the user, - the same holds recursively for all the directories from the root to the file. This query can be neatly expressed as a $\CTP$ path formula: $$\mathtt{/(child::dir[@_{(user,read)}=1][@_{(user,execute)}=1])^*/ child::file[@_{(user, read)}=1]}.$$ Additionally, $\CTP$ can express non-trivial constraints over the tree representing the file system. For instance, the formula $$\mathtt{//self::file[@_{(other,read)}=1]} \rightarrow \mathtt{/(child::dir[@_{(other,read)}=1]} \mathtt{[@_{(other,execute)}=1])^*/ *} $$ imposes the constraint that for all files which are readable it holds that every directory on the path from the root to this file must be both readable and executable by others. It is known that the conditional axis cannot be expressed by tree patterns or in Core XPath [@marx:cond05]. On the other hand, the queries from Example \[ex:cond\] can be expressed in the positive forward fragment of Regular XPath, which is more expressive than $\CTP$. However, we believe this additional expressive power leads to increase in the complexity (of containment) than for conditional tree patterns. Related Work ------------ The complexity of tree patterns is studied in a number of papers and since [@neve:comp06] a virtually complete picture exists for the complexity of the containment problem for positive fragments of XPath. Miklau and Suciu [@mikl:cont04] show that the complexity of the containment problem for $\TP$ with filters, wildcard and descendant is $\coNP$-complete. Containment and equivalence for fragments of $\TP$ were studied before. The most interesting result is that containment for $\TP$ without the wildcard is in $\PTIME$ [@amer:tree02]. Our conditional tree patterns are a first order fragment of conjunctive regular path queries. Calvanese et al. [@calv:cont00] show that the complexity of containment of these queries is $\EXPSPACE$-complete, but these are interpreted on general graph models. Conditional XPath [@marx:cond05] and conditional tree patterns are closely related to branching time temporal logic $\CTL$ [@clar:auto86]. The conditional child axis and the strict until connective are interdefinable. Kupferman and Vardi [@kupf:auto00] show that the containment problem for $\ECTL$, which is the restriction of $\CTL$ to formulas having only the $\exists$ path quantifier in front of them, is a $\PSPACE$-complete problem. The positive $\ECTL$-fragment without until was also studied by Miklau and Suciu [@mikl:cont04]. They show that the containment problem for this fragment is equivalent to the $\TP$-containment problem and thus also $\coNP$-complete. This paper studies the containment problem without presence of schema information. A number of complexity results for containment in $\TP$ w.r.t DTDs are given in [@neve:comp06]. In particular, containment for $\TP$ with filters, the wildcard and descendent w.r.t DTD is $\EXPTIME$-complete. This hardness result together with an $\EXPTIME$ upper bound for Conditional XPath [@marx:cond05], which contains $\CTP$, gives us $\EXPTIME$-completeness of containment for $\CTP$ in the presence of DTDs. Preliminaries {#sec:prelim} ============= We review the basic definitions of XML trees, Regular XPath and its semantics. Then we present Tree Patterns and Conditional Tree Patterns as sublanguages of Regular XPath. Tree patterns have an alternative semantics in terms of embeddings [@mikl:cont04]. We give such an “embedding semantics” for Conditional Tree Patterns using “until-simulations” in Section \[sec:expr\]. Trees ----- We work with node-labeled unranked finite trees, where the node labels are elements of an infinite set $\Sigma$. Formally, a tree over $\Sigma$ is a tuple $(N, E, r, \rho)$, where $N$, the set of nodes of the tree, is a prefix closed set of finite sequences of natural numbers, $E=$ $\{ (\langle n_1, \dots, n_k\rangle, \langle n_1, \dots, n_k, n_{k+1}\rangle)$ $~\mid ~$ $ \langle n_1, \dots, n_{k+1}\rangle \in N\}$ is the edge or child relation, $r$ is the root of the tree, that is the empty sequence, and $\rho$ is the function assigning to each node in $N$ a finite subset of $\Sigma$. We refer a tree over $\Sigma$ just as a tree if $\Sigma$ is clear from the context. Trees in which $\rho(\cdot)$ is always a singleton are called *single-labeled* or *XML trees*. Trees without this restriction are called *multi-labeled trees*. We denote by $E^+$ the descendant relation, which is the transitive closure of the edge relation $E$, and by $E^*$ the reflexive and transitive closure of $E$, and by $E(x)$ the set of all children of the node $x$. A node $n$ is a *leaf* if $E(n)$ is empty. A *path* from a node $n$ to a node $m$ is a sequence of nodes $n=n_0, \dots, n_k=m$, with $k>0$, such that for each $i\leq k$, $n_{i+1} \in E(n_i)$. We call a *branch* any maximal path starting from the root. If $mE^+n$, $m$ is called an *ancestor* of $n$, and if $mEn$, $m$ is called the *parent* of $n$. If $n$ is in $N$, by $T.n$ we denote the subtree of $T$ rooted at $n$. A *pointed tree* is a pair $T,n$ where $T$ is a tree and $n$ is a node of $T$. The *height* of a pointed tree $T,n$ is the length of the longest path in $T.n$. Given two trees $T_1=(N_1,E_1, r_1, \rho_1)$ and $T_2=(N_2, E_2, r_2, \rho_2)$ such that $N_1$ and $N_2$ are disjoint, we define the result of *fusion* of $T_1$ and $T_2$, denoted as $T_1\oplus T_2$, as the tree obtained by joining the trees $T_1$ and $T_2$ without the roots under a new common root labeled by the union of the labels of the roots of $T_1$ and $T_2$. Formally, $T_1\oplus T_2$ is the tree $T=(N,E, r, \rho)$, where $N= (N_1\setminus\{r_1\})\cup (N_2\setminus \{r_2\})\cup \{r\}$, $E=(E_1\setminus \{\tup{r_1,n}\mid n\in N_1\})\cup (E_2\setminus \{\tup{r_2, n}\mid n\in N_2\})\cup \{\tup{r,n}\mid \tup{r_1, n}\in E_1 \text{ or }\tup{r_2, n}\in E_2\}$ and $\rho(n) = \begin{cases} \rho_1(r_1)\cup \rho_2(r_2) & \text{if } n=r, \\ \rho_1(n) & \text{if } n\in N_1\setminus \{r_1\}, \\ \rho_2(n) & \text{if } n\in N_2\setminus \{r_2\}. \end{cases} $ XPath and Tree Patterns ----------------------- We define Tree Patterns and Conditional Tree Patterns as sublanguages of Regular XPath [@cate:expr06]. Let $\Sigma$ be an infinite domain of labels. Forward Regular XPath, $\RXP$ for short, consists of node formulas ${\varphi}$ and path formulas $\alpha$ which are defined by the following grammar ------------- ------- --------------------------------------------------------------------------------------------------------------------------------- ${\varphi}$ $::=$ $p \mid \top \mid \lnot {\varphi}\mid{\varphi}\lor {\varphi}\mid{\varphi}\land {\varphi}\mid \langle \alpha \rangle {\varphi}$ $\alpha$ $::=$ $ \downarrow\ \mid\ ?{\varphi}\mid \alpha ; \alpha \mid \alpha \cup \alpha \mid \alpha^* $, ------------- ------- --------------------------------------------------------------------------------------------------------------------------------- where $p \in \Sigma$. We will use $\alpha^+$ as an abbreviation of $\alpha;\alpha^*$. For the semantics of $\RXP$, given a tree $T=(N, E, r, \rho)$ over $\Sigma$, the relation $\sem{\alpha}_T\subseteq N\times N$ for a path expression $\alpha$ and the satisfaction relation $\models$ between pointed trees and node formulas are inductively defined as follows: - $\sem{\downarrow}_T = E$, - $\sem{ ? {\varphi}}_T = \{(n,n)\in N\times N\mid\, T,n \models {\varphi}\}$, - $\sem{ \alpha;\beta }_T = \sem{\alpha}_T \circ \sem{\beta}_T$, - $\sem{ \alpha\cup\beta }_T = \sem{\alpha}_T \cup \sem{\beta}_T$, - $\sem{\alpha^\ast}_T= (\sem{\alpha}_T)^\ast $, and - $T,n\models \top$, - $T,n\models p$ iff $p\in \rho(n)$, - $T,n\models \lnot{\varphi}$ iff $T,n\not\models{\varphi}$, - $T,n\models {\varphi}\land\psi$ iff $T,n\models{\varphi}$ and $T,n\models\psi$, - $T,n\models {\varphi}\lor \psi$ iff $T,n\models{\varphi}$ or $T,n\models\psi$, - $T,n\models \ch{\alpha}{\varphi}$ iff there is a node $m$ with $(n,m)\in \sem{ \alpha}_T$ and $T,m\models{\varphi}$. Sometimes we will write $T\models {\varphi}$ to denote $T,r\models {\varphi}$. If the latter holds, we say that $T$ is a *model* of ${\varphi}$. ### (Conditional) Tree Patterns {#sec:cond-tree-patt .unnumbered} Tree patterns are the conjunctive fragment of $\RXP$ without unions of paths and with a strongly restricted Kleene star operation. Node formulas correspond to Boolean tree patterns, and path formulas to tree patterns with one output node. In the following definitions we again let $\Sigma$ be an infinite domain of labels. We define Tree Patterns by restricting the syntax of $\RXP$ as follows: Tree Patterns ($\TP$) consist of node formulas ${\varphi}$ and path formulas $\alpha$ defined by the following grammar: ------------- ------- ----------------------------------------------------------------------------------- ${\varphi}$ $::=$ $p \mid \top \mid {\varphi}\land {\varphi}\mid \langle \alpha \rangle {\varphi}$ $\alpha$ $::=$ $ \downarrow\ \mid\ ?{\varphi}\mid \alpha ; \alpha \mid \ \downarrow^+, $ ------------- ------- ----------------------------------------------------------------------------------- where $p \in \Sigma$. For example, the tree pattern from Figure \[fig:tp\] can be written as the path formula $\alpha = ?(p \land \child{t}); \downarrow^+; ?(s\land \child{v})$. Conditional Tree Patterns are also defined by restricting the syntax of $\RXP$, where we allow conditional descendant paths. Conditional Tree Patterns ($\CTP$) consist of node formulas ${\varphi}$ and path formulas defined by the following grammar: ------------- ------- ----------------------------------------------------------------------------------- ${\varphi}$ $::=$ $p \mid \top \mid {\varphi}\land {\varphi}\mid \langle \alpha \rangle {\varphi}$ $\alpha$ $::=$ $ \downarrow\ \mid\ ?{\varphi}\mid \alpha ; \alpha \mid (\downarrow;?{\varphi})^*;\downarrow\enspace$, ------------- ------- ----------------------------------------------------------------------------------- where $p \in \Sigma$. The tree from Figure \[fig:CTP1\] can be written as the path formula $$?(p \land \child{t}); \upath{r}; ?(s\land \tup{ \upath{(a \land \tup{\upath{a}}b)}}v ).$$ Note that the node formula $\langle (\downarrow;?{\varphi})^*;\downarrow \rangle \psi$ is exactly the strict until $\exists U(\psi,{\varphi})$ from branching time temporal logic $\CTL$ [@kupf:auto00]. We will abbreviate this formula simply as $\until{{\varphi}}{\psi}$. Because of the equivalences $\langle \alpha;\beta\rangle{\varphi}\equiv \langle\alpha\rangle \langle\beta\rangle{\varphi}$ and $\langle?{\varphi}\rangle\psi\equiv {\varphi}\wedge\psi$, $\CTP$ node formulas can be given the following equivalent definition: ------------- ------- -------------------------------------------------------------------------------------------------------------------------------------------------------------------- ${\varphi}$ $::=$ $p \mid \top \mid {\varphi}\land {\varphi}\mid \langle \downarrow \rangle {\varphi}\mid \langle \downarrow^+ \rangle {\varphi}\mid \until{{\varphi}}{{\varphi}}$. ------------- ------- -------------------------------------------------------------------------------------------------------------------------------------------------------------------- Even though the syntax is slightly different, $\CXP$ is the conjunctive forward only fragment of Conditional XPath [@marx:cond05]. ### Expansions {#expansions .unnumbered} We consider two expansions of $\TP$ and $\CTP$, with negated labels and with disjunction in node formulas together with union in paths. Negated labels is a restricted type of negation where only the construct $\neg p$ is allowed in the node formulas. We denote expansions of the language $L$ with one or two of these features by $L^S$ for $S\subseteq \{\neg,\vee \}$. ### Query evaluation {#query-evaluation .unnumbered} In [@gott:effi05] it was shown that the query evaluation problem for Core XPath is $\PTIME$-complete in the combined complexity. As noted in [@marx:cond05], using results on model checking for Propositional Dynamic Logic, this can be extended to Regular XPath, and thus to all our defined fragments. Let $T$ be a tree, $n_1, n_2$ nodes in $T$, and $\alpha$ a Regular XPath path formula. The problem whether $(n_1, n_2)\in \sem{\alpha}_T$ is decidable in time $O(|T|\times |\alpha|)$ with $|T|$ the size of the tree and $|\alpha|$ the size of the formula. Containment ----------- As we are considering two kinds of expressions, path and node expressions, we have different notions of containment. Let $\varphi$ and $\psi$ be two $\RXP$-node formulas. We say that $\varphi$ is *contained* in $\psi$, notation ${\varphi}\subseteq \psi$, if for every $T,n$, $T,n\models\varphi$ implies $T,n\models\psi$. Let $\alpha$ and $\beta$ be two $\RXP$-path formulas. We say that - $\alpha$ is *contained as a binary query* in $\beta$, denoted $\alpha \subseteq^2 \beta$, if for any tree $T$, $\sem{\alpha}_T \subseteq \sem{\beta}_T$, - $\alpha$ is *contained as a unary query* in $\beta$, denoted , if for any tree $T$ with root $r$, and any node $n$, $(r, n) \in \sem{\alpha}_T$ implies $(r,n) \in\sem{\beta}_T$. Containment over single-labeled trees is denoted by $\subseteq$, and containment over multi-labeled models by $\mlcont$. Luckily, these three notions are closely related, and containment of path formulas can be reduced to containment of node formulas and vice versa (cf. also [@neve:comp06; @mikl:cont04]). \[prop:output\] Let $\alpha$ and $\beta$ be two $\RXP$-path formulas, ${\varphi}$ and $\psi$ $\RXP$-node formulas, in which negation is restricted to labels only. 1. $\alpha \subseteq^2 \beta$ iff $\alpha \subseteq^1 \beta$, 2. Let $p$ be a label not occurring in $\alpha$ or in $\beta$. Then, $\alpha \subseteq^2 \beta$ iff $\langle \alpha \rangle\langle \downarrow \rangle p \subseteq \langle \beta \rangle\langle \downarrow \rangle p$, 3. ${\varphi}\subseteq \psi$ iff $?{\varphi}\subseteq^2 ?\psi$. All the above items also hold for the case of multi-labeled trees. \(i) ($\Rightarrow$) Let $T=(N, E, r, \rho)$ be a tree such that $(r, n)\in \sem{\alpha}_T$. By the assumption, we have $\sem{\alpha}_T\subseteq \sem{\beta}_T$. This implies that $(r, n)\in \sem{\beta}_T$, which was required to show. ($\Leftarrow$) Let $T=(N, E, r, \rho)$ be a tree, $n_1$ and $n_2$ in $N$ such that $(n_1, n_2)\in \sem{\alpha}_T$. Since $\alpha$ is from the forward fragment of Regular XPath, we have that $(n_1, n_2)\in \sem{\alpha}_{T.n_1}$. Using the assumption, we have $(n_1, n_2)\in \sem{\beta}_{T.n_1}$. Then by monotonicity, we obtain $(n_1, n_2)\in \sem{\beta}_T$, which was desired. \(ii) ($\Rightarrow$) Suppose $\alpha \subseteq^2 \beta$, and let $T, n \models \tup{\alpha}\tup{\downarrow} p$. Thus there is a descendant $m$ of $n$ in $T$ such that $(n,m) \in \sem{\alpha}_T$ and $T, m \models \tup{\downarrow}p$. By the hypothesis we therefore have $(n,m) \in \sem{\beta}_T$ and thus $T, n \models \tup{\beta}\tup{\downarrow}p$. ($\Leftarrow$) Assume $\tup{\alpha} \tup{\downarrow} p \subseteq \tup{\beta} \tup{\downarrow} p$. Consider a tree $T=(N, E, r, \rho)$ and a pair $(n,m) \in \sem{\alpha}_T$. Since $p$ does not occur in $\alpha$, we may assume that $T, n\not\models p$ for every node $n\in N$. We then define the tree $T'=(N', E', r, \rho')$, where - $N':=N\cup \{x\}$, - $E':=E'\cup \{\tup{m,x}\}$, - $\rho'(y):= \left\{ \begin{array}{l l} p & \quad \text{if\ } y= x,\\ \rho(y) & \quad \text{otherwise.} \end{array} \right. $ By definition of $T'$, we have $T', m\models \tup{\downarrow} p$ and $(n, m)\in \sem{\alpha}_{T'}$. Hence, $T', n\models \tup{\alpha}\tup{\downarrow}p$. By the assumption, it implies that $T', n\models \tup{\beta}\tup{\downarrow}p$. Since $p$ holds at $x$ only, and $x$ is a child of $m$, we have that $(n, m)\in \sem{\beta}_{T'}$. By definition of $T'$, the path between $n$ and $m$ is also in $T$. Thus, $(n, m)\in \sem{\beta}_T$, as desired. Item (iii) easily follows from the definitions. In light of Proposition \[prop:output\], from now on we consider the containment problem of node formulas only. We now give an interesting example of $\CTP$ containment. Let us consider the $\CXP$ node formulas $$\begin{aligned} \label{eq:2} {\varphi}&=&\until{(\until{b}{d})}{\until{a}{b}} \ \text{and}\\ \psi &=&\until{(\until{c}{d})}{\until{a}{b}}.\end{aligned}$$ Although it is hard to see from the first glance, ${\varphi}\subseteq \psi$ holds. Indeed, in every model $T$ of ${\varphi}$ either there is a direct child of the root where $\until{a}{b}$ holds, or there is a path $r=v_1,\ldots, v_n$ in $T$ to the node $v_n$ where $\until{a}{b}$ holds and $\until{b}{d}$ holds in every node $v_i$ ($1< i< n$) on the path. In the first case, $\psi$ holds at the root since $\until{a}{b}$ holds at the direct child. Now let’s consider the second case. Let $j\in \{2,\ldots,n-1\}$ be the least number with the property that $v_j$ has a non-empty $b$-path to the $d$-descendant. If there is no such number, then $\until{c}{d}$ holds at every $v_i, 1< i<n$ (as each of them has a direct $d$-child) and, thus, $T,r\models \psi$. Assume there is such a $j$. Since $v_j$ has a $b$-node as a child, we have that $T,v_j\models \until{a}{b}$. Moreover, $\until{c}{d}$ holds at each $v_i$ for $i< j$ since $v_i$ has a $d$-node as a child. Thus in this case we obtain that $T,r\models \psi$ holds too. Expressivity {#sec:expr} ============ We extend the semantics of $\TP$ given by embeddings of queries into trees to $\CTP$. Instead of embeddings we need a simulation known from temporal logic. Interpreting Conditional Tree Patterns by simulations ----------------------------------------------------- The semantics of conditional tree patterns can be defined using simulations developed for LTL [@blac:moda01]. These simulations generalize the embeddings for tree patterns from [@amer:tree02; @mikl:cont04] with an additional clause for checking the labels on the edges. We start with defining the tree pattern analogues of $\CTP$ node and path expressions. A *conditional tree pattern* is a finite tree $(N, E , r, \bar{o}, \rho_N, \rho_E)$ with labeled nodes and edges, where $N$ is the set of nodes of the tree, $E\subseteq N\times N$ is the set of edges, $r$ is the root of the tree, $\bar{o}$ is a $k$-tuple ($k> 0$) of output nodes, $\rho_N$ is the function assigning to each node in $N$ a finite set of labels from $\Sigma$ and $\rho_E$ is the function assigning to each pair in $E$ either $\downarrow$ or a Boolean conditional tree pattern.\ A *Boolean conditional tree pattern* is a conditional tree pattern with a single output node which equals to the root.\ A conditional tree patterns is said to have *multiple output nodes* if the number $k=|\bar{o}| $ is greater than 1.\ A *tree pattern* is a conditional tree pattern whose edges are only labeled by $\downarrow$ or $\top$. To be consistent with the pictorial representation of $\TP$, in $\CTP$ patterns an edge labeled with $\downarrow$ is drawn as a single line, while an edge labeled with a $\CTP$ node formula is drawn as a double line with a $\CTP$ pattern as the label (e.g. as in Figure \[fig:CTP1\]). The output nodes have the square shape. $\CTP$ node and path expressions can be translated into (Boolean) conditional tree patterns with one output node and vice-versa. The translations are given in Appendix \[app:tra\]. We denote the equivalent (Boolean) conditional tree pattern of a $\CTP$ path or node expression $\alpha$ or ${\varphi}$ by $\ct(\alpha)$ and $\ct({\varphi})$, respectively. Next we generalize the notion of $\TP$-embeddings to $\CTP$-simulations. \[def:sim\] Let $T=(N, E, r, \bar{o},\rho_N, \rho_E)$ be a conditional tree pattern as in the previous definition and $T'=(N', E', r', \rho')$ a tree. A total function $f: N \to N'$ is called a *simulation* from the pattern $T$ into the pointed tree $T',r'$ if it satisfies the following properties: root preserving : $f(r)=r'$; label preserving : if $p \in \rho_N(n)$, then $p \in \rho'(f(n))$; child edge preserving : if $nEn'$ and $\rho_E(n,n')= '\downarrow'$, then $f(n)Ef(n')$; conditional edge simulation : if $nEn'$ and $\rho_E(n,n')$ is not equal to $\downarrow$, then $f(n)E'^+f(n')$ and for every $x$ such that $f(n)E'^+x E'^+f(n')$ there is a simulation from the Boolean conditional tree pattern $\rho_E(n,n')$ into $T'.x$ (the subtree of $T'$ rooted at $x$). When the pattern is a tree pattern, $\downarrow$ and $\top$ are the only labels on edges. For the label $\top$, the ’conditional edge simulation’ clause trivializes to checking that $f$ is an embedding for the descendant edges. Simulations for tree patterns are thus equivalent to tree pattern embeddings [@amer:tree02; @mikl:cont04]. The next theorem states that simulations can be used to evaluate conditional tree patterns. \[prop:simc\] Let ${\varphi}$ and $\alpha$ be a $\CXP$ node and path expression, respectively. Let $T$ be a tree and $n$ a node in $T$. 1. $T,n \models {\varphi}$ if and only if there is a simulation from $\ct({\varphi})$ into $T.n$. 2. $(n,n')\in \sem{\alpha}_T$ if and only if there is a simulation from $\ct(\alpha)$ into $T.n$ which relates the output node of $\ct(\alpha)$ to node $n'$. 3. Items (i) and (ii) also hold when $T$ is an infinite tree. The proof is by mutual induction on the node and path expressions. \[rem1\] The notions of (conditional) tree pattern and embedding for tree patterns and simulation for conditional tree patterns are easily extended to work for expansions of these languages with negated labels (denoted by $\TP^{\neg}$ and $\CTP^{\neg}$, respectively). For (conditional) tree patterns, add a second node labelling function $\rho_N^{\neg}(\cdot)$ which also assigns each node a finite set of labels. These are interpreted as the labels which are false at the node. For the embeddings and simulations, in Definition \[def:sim\] we add a clause stating that also negated labels are preserved: negated label preserving : If $ p\in \rho_N^{\neg}(n)$, then $p\not\in \rho'(f(n))$. With these modifications, Theorem \[prop:simc\] also holds for $\TP^{\neg}$ and $\CTP^{\neg}$. Expressivity characterization ----------------------------- In this section we give expressivity characterization for $\CTP$ similar to the one for various fragments of XPath in [@bene:stru05]. The exact logical characterization allows one to compare different fragments of (Regular) XPath and derive non-trivial closure properties such as closure under intersection. We show that $\CTP$ path formulas correspond to a natural fragment of first order logic ($\mathbf{FO}$) and are closed under intersections. For ${\varphi}$ a formula, let $\desphi{{\varphi}}(x,y)$ be an abbreviation of the “until” formula $desc(x,y) \wedge \forall z(desc(x,z)\wedge desc(z,y)\rightarrow {\varphi}(z))$. Let $\FOCTP$ be the fragment of first order logic built up from the binary relations $child$ and $desc$, label predicates $p(x)$ for each label $p\in \Sigma$ and equality $'{=}'$, by closing under $\wedge,\ \vee$ and $\exists$ as well as under the rule: $$\begin{aligned} \text{if } {\varphi}(x)\in \FOCTP, \text{ then } \desphi{{\varphi}}(x,y) \in \FOCTP. \label{crule}\end{aligned}$$ We use $\FOCTP(c,s)$ to denote $\FOCTP$ formulas with exactly two variables $c$ and $s$ free. Following [@bene:stru05] we restrict the $\mathbf{FO}$ fragment to its downward fragment $\FOCTPd$ by requiring that every bound variable as well as $s$ is syntactically restricted to be a descendant of $c$ or equal to $c$. Note that $\FOCTPd$ without the rule is the logic $\FOBFKd$ from [@bene:stru05], and shown there to be equivalent to unions of tree patterns. Even though $\FOCTPd$ does not contain negation, not every formula is satisfiable (e.g. $desc(x,x)$). We can now characterize conditional tree patterns in terms of $ \FOCTPd$. \[thm:exp1\] The following languages are equivalent in expressive power: - unions of the false symbol and $\CTP$ paths which can have disjunctions in the node formulas - $ \FOCTPd$. A standard translation $TR_{xy}(\cdot)$ turns any $\CTP$ path formula $\alpha$ into an equivalent formula in $\FOCTPd$. The translation is essentially the semantics of path and node formulas written in the first order language. ---------------------------------------- --- ---------------------------------------------------------- $TR_{xy}(\emptyset)$ = $child(x,y)\land x=y$ $TR_{xy}(\downarrow)$ = $child(x,y)$ $TR_{xy}(?{\varphi})$ = $ x=y \land TR_x({\varphi})$ $TR_{xy}(\alpha_1; \alpha_2)$ = $\exists z. (TR_{xz}(\alpha_1)\land TR_{zy}(\alpha_2))$, where $z$ is a fresh variable $TR_{xy}(\untilpath)$ = $\desphi{TR_z({\varphi})}(x,y)$ $TR_{xy}(\alpha_1\cup \alpha_2)$ = $TR_{xy}(\alpha_1)\lor TR_{xy}(\alpha_2)$ $TR_{x}(p)$ = $p(x)$ $TR_{x}(\top)$ = $x=x$ $TR_{x}({\varphi}_1\land {\varphi}_2)$ = $TR_x({\varphi}_1)\land TR_x({\varphi}_2)$ $TR_{x}({\varphi}_1\lor {\varphi}_2)$ = $TR_x({\varphi}_1) \lor TR_x({\varphi}_2)$ $TR_{x}(\tup{\alpha}{\varphi})$ = $\exists y. (TR_{xy}(\alpha)\land TR_y({\varphi}))$, where $y$ is a fresh variable. ---------------------------------------- --- ---------------------------------------------------------- By definition, $TR_{cs}(\alpha)$ is a formula in $ \FOCTPd$ and equivalent to $\alpha$. For the other direction, let $\theta\in \FOCTPd$. We use our earlier notation $\desphi{\psi}(x,y)$ for the “until” formulas. First we introduce two special formulas $\top(x)$ and $\bot(x)$ which stand for $x=x$ and $desc(x,x)$, respectively. We will modify $\theta$ in several steps. First we replace $child(x,y)$ and $desc(x,y)$ by the equivalent $\desphi{\bot}(x,y)$ and $\desphi{\top}(x,y)$, respectively. Then eliminate all equalities by renaming variables. Bring all existential quantifiers inside $\theta$ to the front, bring the body into disjunctive normal form and distribute the disjunctions over the quantifiers. We then end up with a disjunction of formulas of the form $\exists\bar{x}{\varphi}(c,s)$, with ${\varphi}$ a conjunction of formulas $\desphi{\psi}(x,y)$, $p(x)$, $\top(x)$ and $\bot(x)$. If ${\varphi}$ contains $\bot$ replace $\exists\bar{x}{\varphi}(c,s)$ with $\bot$. As our target language is closed under unions and $\bot$, we only need to translate the “conjunctive queries” $\exists\bar{x}{\varphi}(c,s)$. Consider the graph of the variables in ${\varphi}$ in which two variables $x,y$ are related if ${\varphi}$ contains an atom $\desphi{\psi}(x,y)$. If the graph is cyclic, it cannot be satisfied on a tree and $\exists\bar{x}{\varphi}(c,s)$ is equivalent to $\bot$. If it is a tree, $c$ will be the root. Assuming that we can rewrite all $\psi$ in the atoms $\desphi{\psi}(x,y)$, we can rewrite it as a conditional tree pattern. If $\psi$ is a boolean combination of $p(z)$, $\top(z)$ and $\bot(z)$ atoms, this is not hard: Simply bring it into disjunctive normal form, and remove each conjunct containing $\bot$. If the result is not empty, then translate to a union of (trivial) conditional tree patterns. If it is empty, translate $\desphi{\psi}(x,y)$ to a child step. If $\psi$ contains subformulas of the form $\desphi{{\varphi}}(x,y)$, we apply the current procedure to it. Thus assume that the variable graph is a directed acyclic graph. We eliminate undirected cycles step by step. The length of the cycle equals the number of variables in it. Consider that the graph contains two paths $\pi_1$ and $\pi_2$ both going from $x$ to $y$ and without other common variables. We show that this subgraph is equivalent to a union of subgraphs with no or smaller cycles. In the simplest case, both paths are of length 1 and thus consist of a $\desphi{\psi}(x,y)$ atom. But then we can use the following equivalence to remove the cycle: $$\desphi{{\varphi}}(x,y)\wedge \desphi{\psi}(x,y) \equiv \desphi{\psi\wedge\psi}(x,y)$$ So assume the cycle looks like Figure \[fig:paths\]. ![undirected cycle in ${\varphi}$[]{data-label="fig:paths"}](paths1a) If this is satisfied on a tree $(N,E, r,\rho)$ with assignment $g$, there are three possibilities: $g(z_1)=g(z_2)$, $g(z_1)E^+ g(z_2)$ or $g(z_2)E^+g(z_1)$. In each case our original formula is equivalent to one with a smaller cycle. The three possibilities are when $g(z_1)=g(z_2)$ : $\desphi{{\varphi}\wedge\psi}(x,z_1) \wedge z_1\pi_1'y \wedge z_1\pi_2'y$ when $g(z_1)E^+ g(z_2)$ : $ \desphi{{\varphi}\wedge\psi}(x,z_1) \wedge \psi(z_1) \wedge z_1\pi_1'y \wedge \desphi{\psi}(z_1,z_2)\wedge z_2\pi_2'y$ when $g(z_2)E^+g(z_1)$ : $ \desphi{{\varphi}\wedge\psi}(x,z_2) \wedge {\varphi}(z_2) \wedge z_2\pi_2'y \wedge \desphi{{\varphi}}(z_2,z_1)\wedge z_1\pi_1'y$. In the first case the length of the cycle decreased by two, in the two other cases by one. Thus their disjunction is equivalent to the formula of Figure \[fig:paths\]. We replace that formula by this disjunction, bring the result in disjunctive normal form and distribute the disjuncts out. We again have a disjunction of “conjunctive queries”. As the new cycles are smaller, this procedure will terminate, and results in a (big) disjunction of trees. An important consequence of this result is that $\CTP$ patterns are closed under intersection: \[thm:exp3\] The intersection of two $\CTP$ paths is equivalent to $\bot$ or a union of $\CTP$ paths. Theorem \[thm:exp1\] together with the translation of Regular XPath into $\mathbf{FO}^*(c,s)$ from [@cate:expr06] implies that every union of $\CTP$ patterns with disjunctions is in the intersection of first order logic and positive downward $\mathbf{FO}^*(c,s)$. It is an intriguing open problem whether the converse also holds. Containment {#sec:cont} =========== Before we determine the exact complexity of the containment problem for $\CTP$ and expansions we prove a number of reductions. Most are generalizations from $\TP$ to $\CTP$. The key new result is the encoding of negated labels in $\CTP$. Containment Preliminaries ------------------------- The following reductions will be used later in our upper and lower bound proofs. ### Multiple output nodes {#sec:mult-outp-nodes .unnumbered} The main difference between tree patterns and their XPath formulation is that tree patterns can have multiple output nodes. Kimelfeld and Sagiv [@kime:revi08] (Proposition 5.2) show that for the $\TP$ containment problem the number of output nodes is not important: the problem can be $\PTIME$ reduced to a containment problem of Boolean $\TP$s. This is achieved, given ${\varphi}, \psi\in \TP$, by adding a child labeled with a new label $a_i$ to every output node $X_i$ in both ${\varphi}$ and $\psi$. Second, to every leaf that is not an output node or a newly added node, we add a child labeled with $\top$. The same result, using the same argument, holds for $\CTP$. \[prop:boolean\] Let $S\subseteq\{\vee, \neg\}$. For $\CTP^S$ patterns with multiple output nodes $\varphi, \psi$ there are $\PTIME$ computable Boolean $\CTP^S$, ${\varphi}', \psi'$ such that ${\varphi}\subseteq \psi\ \mbox{iff}\ {\varphi}'\subseteq \psi'.$ The same holds for multi-labeled trees. ### Disjunctions in the consequent {#disjunctions-in-the-consequent .unnumbered} We now show that the containment problem for both unions of $\CTP^{\neg}$ and unions of $\TP^{\neg}$ can be reduced to containments without unions. This is useful in lower bound proofs, as we can use a union in the consequent to express multiple constraints. The proof of the following proposition is a slight modification of the proof for $\TP$ in [@mikl:cont04]. ------------------------------------------------------------------------------------------------------------------------------------- ------------------------------------------------------------------------------------------------------------------------------------- ![Patterns ${\varphi}'$ and $\psi$ from Proposition \[prop:mik\].[]{data-label="fig12"}](figuredisjunct1 "fig:"){width="20.00000%"} ![Patterns ${\varphi}'$ and $\psi$ from Proposition \[prop:mik\].[]{data-label="fig12"}](figuredisjunct2 "fig:"){width="20.00000%"} \[fig12\](a) Pattern ${\varphi}'$ \[fig12\](b) Pattern $\psi$ ------------------------------------------------------------------------------------------------------------------------------------- ------------------------------------------------------------------------------------------------------------------------------------- ![Model for pattern ${\varphi}'$ from Proposition \[prop:mik\].[]{data-label="fig:3"}](figuredisjunct3){width="25.00000%"} \[prop:mik\] Let ${\varphi}$ be a $\CXP^{\neg}$ ($\TP^{\neg}$) formula and $\Delta$ a finite set of $\CXP^{\neg}$ (resp. $\TP^{\neg}$) formulas. Then there are $\PTIME$ computable $\CXP^{\neg}$ ($\TP^{\neg}$) formulas ${\varphi}'$ and $\psi$ such that ${\varphi}\subseteq \bigvee \Delta$ iff ${\varphi}' \subseteq \psi$.\ If ${\varphi}$ and $\Delta$ are in $\CTP$ ($\TP$), ${\varphi}'$ and $\psi$ are in $\CTP$ ($\TP$) as well.\ The same holds for containment over multi-labeled trees. First, for the case of $\TP^{\neg}$, the proof from [@mikl:cont04] can be readily applied here (using embeddings which also preserve negated labels, cf Remark \[rem1\]). Now we prove the proposition for $\CTP^{\neg}$. For simplicity, we use the same letter to denote a $\CXP^{\neg}$ formula and its corresponding tree pattern representation. Assume $\Delta$ is $\{\delta_1, \dots, \delta_k\}$. If $\iota$ is a $\CXP^{\neg}$ pattern, by $\iota^a$ we denote the $\CTP^{\neg}$ pattern defined as having the root labeled by $a$, whose unique child is the root of $\iota$. Consider new labels $a$ and $b$ occurring neither in ${\varphi}$ nor $\Delta$. Let $\gamma=\bigwedge_{0<\ell \leq k} \delta^b_\ell$. Note that $\gamma$ is consistent if every $\delta_\ell$ is consistent. Now define ${\varphi}'$ as the $\CXP^{\neg}$ tree pattern in Figure \[fig12\](a), and $\psi$ as the $\CXP^{\neg}$ tree pattern in Figure \[fig12\](b). We verify that ${\varphi}\subseteq \bigvee \Delta$ iff ${\varphi}' \subseteq \psi$. Without loss of generality, we assume that each of ${\varphi}, \delta_1, \dots, \delta_k$ is consistent. For the direction from left to right, assume ${\varphi}\subseteq \bigvee \Delta$ and $T, n \models {\varphi}'$ for an arbitrary $T,n$. This means that there is a path $n=m_0..... m_{2k-1}$ of nodes in $T$ which are labeled with $a$ and such that: - for every $0< \ell \leq k$: $T, m_i \models (\delta_\ell^b)^a$, with $i \in \{1, \dots, k-1\} \cup \{k+1, \dots, 2k-1\}$, - $T, m_k \models ({\varphi}^b)^a$. By the assumption, it follows that there is an index $j$ such that $T, m_k \models (\delta_j^b)^a$. Then, by Theorem \[prop:simc\], there is a simulation $f'$ from $(\delta_j^b)^a$ into $ T.m_k$. The simulation $f'$ can be extended to a simulation $f$ from $\psi$ into $T,n$ in the obvious way such that $f$ is a simulation from $(\delta^b_i)^a$ into $ T,m_{k-j+i}$ and the $a$-descendant edge in $\psi$ is simulated on the path $m_0, \ldots m_{k-j+1}$. Thus, $T,n\models \psi$. For the other direction, assume ${\varphi}' \subseteq \psi$ and $T,n \models \varphi$ for an arbitrary $T$ rooted at $n$. As all $ \delta_\ell$ are consistent, there are trees $T_\ell, n_\ell \models \delta_\ell$, for each $\ell \in \{1, \dots, k\}$. Let $T'$ be the tree created from $T.n$ and the $T_\ell$’s depicted in Figure \[fig:3\] with root $r$. Then $T', r \models {\varphi}'$. Because we assumed ${\varphi}' \subseteq \psi$, then also $T', r \models \psi$, which by Theorem \[prop:simc\] implies the existence of a simulation $g$ from $\psi$ into $ T'.r$. In particular, the span of $k$ $a$-nodes in $\psi$ has to be simulated on an $a$-path of length $k$ in $T'$. This means that for some $j \in \{1, \dots, k\}$, $g$ is a simulation from $\delta_j$ into $T.n$. By Theorem \[prop:simc\] again, $T, n \models \delta_j$, and thus $T, n \models \bigvee \Delta$. Note that if ${\varphi}$ and $\Delta$ are in $\CTP$, then the constructed ${\varphi}'$ and $\psi$ are in $\CTP$ too. ### From single-labeled to multi-labeled trees {#from-single-labeled-to-multi-labeled-trees .unnumbered} We now show that the containment problem over single-labeled trees can be reduced to a containment problem over multi-labeled trees. \[prop:standard\] Let $S\subseteq \{\neg,\vee\}$. Given $\CXP^S$ ($\TP^S$) patterns ${\varphi},\psi$, there are $\PTIME$ computable $\CTP^S$ ($\TP^S$) patterns ${\varphi}',\psi'$ such that ${\varphi}\subseteq\psi$ iff ${\varphi}'\mlcont\psi'$. Let $p_1, \ldots, p_n$ be the labels from $\Sigma$ occurring in ${\varphi}$ or $\psi$. We take ${\varphi}''$ as ${\varphi}$, and $\psi'':=\psi \vee\bigvee_{1\leq i<j\leq n} (p_i\wedge p_j) \vee\bigvee_{1\leq i<j\leq n} \until{\downarrow}{(p_i\wedge p_j)}$. The disjuncts added to $\psi$ ensure that every node of a counterexample is labeled with at most one label. Now we show that ${\varphi}\subseteq\psi$ iff ${\varphi}''\mlcont\psi''$ ($\Rightarrow$) We show the contraposition. Let $T$ be a multi-labeled tree such that $T\models {\varphi}''$ and $T\not\models \psi''$. Then let $T'$ be the tree obtained from $T$ by restricting the labels $\rho'(v):=\rho(v)\cap \{p_1,\ldots , p_n\}$ for every node $v$. Since $T$ does not satisfy $\psi''$, the label of every node in $T'$ contains at most one symbol. Because we did not change the valuation of the labels $p_1, \ldots, p_n$, for each formula $\theta$ constructed from these tags, it holds that $T,v\models \theta$ iff $T',v\models \theta$. From this and the fact that the label of each node of $T'$ is of size at most one, it follows that $T'\models {\varphi}''$ and $T'\not \models \psi''$. In order to make a single-labeled tree out of $T'$-2pt, we add a dummy symbol $q$ in the label of a node if its label in $T'$ is empty. For the resulting tree $T''$ it holds $T''\models {\varphi}$ and $T''\not \models \psi$. ($\Leftarrow$) Let $T$ be a single-labeled tree such that $T\models {\varphi}$ and $T\not\models \psi$. Then this tree can be considered as a multi-labeled tree. Moreover, $T\models {\varphi}''$ and $T\not\models \psi''$ since $T$ does not satisfy any of the disjuncts in $\psi''$. Our transformation does not introduce negations, but only disjunctions in $\psi''$. Thus the proposition is proven for the cases when $S$ contains disjunction, where we take ${\varphi}':={\varphi}''$ and $\psi':=\psi''$. For the remaining cases ($S\subseteq \{\neg\}$), although $\psi''$ contains disjunctions, by Proposition \[prop:mik\], there exist two $\PTIME$ computable $\CTP^S$ ($\TP^S$) patterns ${\varphi}'$ and $\psi'$ such that ${\varphi}''\mlcont\psi''$ iff ${\varphi}'\mlcont\psi'$. This concludes the proof. ### Negated labels {#sec:negated-labels .unnumbered} We now show that $\CTP$ is expressive enough to encode label negation, as far as the containment problem is concerned. The trick of the encoding lies in the fact that two nodes connected by a descendent edge are in either child or descendant of a child relation. Additionally, if we require that neither of these relations occur at the same time, we can faithfully encode label negation. \[prop:neg\] Let ${\varphi}$ and $\psi$ be $\CTP^\neg$ patterns. There are $\PTIME$ computable $\CTP$ patterns ${\varphi}'$ and $\psi'$ such that $${\varphi}\subseteq\psi\ \mbox{iff}\ {\varphi}'\subseteq \psi'.$$ This also holds for containment over multi-labeled trees. Given a containment problem ${\varphi}\subseteq\psi$ in $\CXP^\neg$, we construct an equivalent containment problem ${\varphi}^\bullet\subseteq\psi^\circ$ with ${\varphi}^\bullet$ in $\CTP$ and $\psi^\circ$ a union of $\CTP$ patterns. Applying Proposition \[prop:mik\] then yields the desired result. Let $p_1$, $\neg p_1,\dots,p_n,\neg p_n$ be the labels appearing in ${\varphi}$ and $\psi$ and their negations. Let $s$ be a new label and $\zeta$ be the formula $\ch\downarrow (s\land \bigwedge_i \until{\downarrow}{p_i})$. We define the translation $(\cdot)^\bullet$ inductively: $$\begin{aligned} \top^\bullet &= \zeta\\ (p_i)^\bullet &= \zeta \land \ch\downarrow(s\land\ch\downarrow p_i)\\ (\neg p_i)^\bullet &= \zeta\land\ch\downarrow(s\land\ch\downarrow\until{\downarrow}{p_i})\\ (\theta\land\sigma)^\bullet &= \theta^\bullet\land\sigma^\bullet \\ (\ch\downarrow{\varphi})^\bullet &= \zeta\land\ch\downarrow{\varphi}^\bullet \\ (\until{\varphi}\psi)^\bullet &= \zeta\land\until{{\varphi}^\bullet}{\psi^\bullet}.\end{aligned}$$ The translation is defined in such a way that for any ${\varphi}$, ${\varphi}^\bullet$ implies $\zeta$. Obviously $(\cdot)^\bullet$ is a $\PTIME$ translation. Let $\descstar {\varphi}$ denote ${\varphi}\lor \until\downarrow{\varphi}$. Let $AX$ be the disjunction of the following formulas: $$\begin{aligned} \descstar(s\land \descstar \zeta) \label{ax:1} \\ \bigvee_i \descstar( p_i^\bullet\land(\neg p_i)^\bullet) \label{ax:2}\\ \bigvee_{i\neq j}\descstar(p_i^\bullet \wedge p_j^\bullet) \label{ax:3}.\end{aligned}$$ In our intended models, none of these disjuncts are true at the root. The disjunct is technical, states that no node makes the encodings of both $p$ and $\neg p$ true, and ensures that no node makes the encoding of two different labels true. We define $\psi^\circ = \psi^\bullet\lor AX$. In case of multi-labeled trees the translation is the same, except that we do not include the disjunction in the definition of $\psi^\circ$. Note that although in $\psi^\circ$ we have disjunctions within the scope of a modality, we can (in $\PTIME$) rewrite the formula into an equivalent union of disjunction free formulas. We now show that ${\varphi}\subseteq\psi$ ${\varphi}^\bullet\subseteq\psi^\circ$.\ ($\Leftarrow$) We show the contrapositive. Given a single-labeled tree $T=(N, E, r, l)$ such that $T,r\models {\varphi}$ and $T, r\not\models \psi$, we build a single-labeled tree $T'=(N', E', r', l')$ as follows. The set of nodes $N'$ is the maximal $E'$-connected subset of $N\times\{0,1,2,3\}\times \{p_1, \ldots, p_n, \sharp\}$ which contains the root $r'=(r,0,\sharp)$. The relation $E'$ is defined as follows: $(n',0,\sharp)$ is the parent of $(n,0,\sharp)$ when $n'$ is the parent of $n$ in $T$; $(n,0,\sharp)$ is the parent of $(n,1,\sharp)$; $(n,1, \sharp)$ is the parent of $(n,2,p_i), i=1, \ldots, n$, and $(n, 2, p_i)$ is the parent of $(n,3,p_i), i=1, \ldots, n$. The labeling $l'$ is defined as follows: $s$ only labels nodes of the form $(n,1,\sharp)$. Then, $p_i$ labels $(n,2,p_i)$ if $T,n\models p_i$ and $p_i$ labels $(n,3,p_i)$ if $T,n\not\models p_i$. No other nodes are labeled by $p_i$. In all other cases, we label a node by a fresh label $z$. It is clear from the definition that $T'$ is a single-labeled tree. Note that $T'\not\models AX$ and that $\zeta$ is true at all nodes of type $(n,0,\sharp)$, and only at these nodes. Let $\theta$ be a formula in variables $\{p_1, \ldots, p_n\}$, and $n\in N$. By induction on $\theta$, we show $T,n\models\theta$ iff $T',(n,0,\sharp)\models \theta^\bullet$. - $\theta=\top$. This holds because by construction $\zeta$ is true at all nodes of type $(n, 0, \sharp)$. - $\theta=p$. We have $T, n\models p$ $\Leftrightarrow$ $l(n)=p$ $\Leftrightarrow$ $ l'(n, 2, p)=p$ $\Leftrightarrow$(since $(n, 2, p)$ is the only child of $(n, 1, \sharp)$ labeled with $p$ and $(n, 1, \sharp)$ is labeled by $s$)\ $T', (n, 1, \sharp)\models s\land \child{p}$ $\Leftrightarrow$ (since $(n, 1, \sharp)$ is the only child of $(n,0, \sharp)$ and $\zeta$ is always true at $(n, 0, \sharp)$) $T', (n, 0, \sharp)\models \zeta\land \child{(s\land \child{p})}$ $\Leftrightarrow$ $T', (n,0, \sharp)\models p^\bullet$. - $\theta=\neg p$. We have $T, n \models \neg p$ $\Leftrightarrow$ $l(n)\neq p$ $\Leftrightarrow$ $l'(n, 3, p)=p$ $\Leftrightarrow$ (since $(n, 3, p)$ is the only descendent of $(n, 1, \sharp)$ labeled with $p$ and $(n, 1, \sharp)$ is labeled by $s$)\ $T', (n, 1, \sharp)\models s\land \child{\dEsc{p}}$ $\Leftrightarrow$ (since $(n, 1, \sharp)$ is the only child of $(n,0, \sharp)$ and $\zeta$ is always true at $(n, 0, \sharp)$) $T', (n,0,\sharp)\models \zeta\land \child{(s\land \child{\dEsc{p}})}$ $\Leftrightarrow$ $T', (n, 0 , \sharp)\models (\neg p)^\bullet$. - $\theta={\varphi}_1\land {\varphi}_2$. We have $T, n \models {\varphi}_1\land {\varphi}_2$ $\Leftrightarrow$ $T, n\models {\varphi}_1$ and $T, n \models {\varphi}_2$ $\Leftrightarrow$ (by the induction hypothesis) $T', (n,0, \sharp)\models {\varphi}^\bullet_1$ and $T', (n, 0, \sharp)\models {\varphi}^\bullet_2$ $\Leftrightarrow$ $T', (n, 0, \sharp)\models {\varphi}^\bullet_1 \land {\varphi}^\bullet_2$ $\Leftrightarrow$ $T', (n,0, \sharp)\models ({\varphi}_1\land{\varphi}_2)^\bullet$. - $\theta=\child{{\varphi}}$. We only show the right to left direction. $T', (n, 0, \sharp)\models (\child{{\varphi}})^\bullet$ $\Leftrightarrow$ $T', (n, 0, \sharp)\models \zeta\land \child{{\varphi}^\bullet}$ $\Leftrightarrow$ $\exists m \in N'$ such that $(n, 0, \sharp)E' m$ and $T',m \models {\varphi}^\bullet$. But then $T',m \models \zeta$ because ${\varphi}^\bullet$ implies $\zeta$. Thus $m$ must be of the form $(n',0,\sharp)$ as only these make $\zeta$ true. But then $n'\in N$ and $nEn'$ and we may apply the inductive hypothesis to get $T,n'\models{\varphi}$, and thus $T,n\models \child{{\varphi}}$. - $\theta=\until {\varphi}\psi$. We only show the right to left direction. $T',(n, 0, \sharp)\models (\until{\varphi}\psi)^\bullet$ $ \Leftrightarrow$ $T', (n, 0, \sharp)\models \zeta \land \until{{\varphi}^\bullet}{ \psi^\bullet}$ $ \Leftrightarrow$ $\exists m \in N'$ such that $(n,0, \sharp)E'^+m$ and $T', m\models \psi^\bullet$ and $\forall m'. (n,0, \sharp)E'^+ m' E'^+ m$ it holds $T', m'\models {\varphi}^\bullet$. Now, because ${\varphi}^\bullet$ implies $\zeta$ for all ${\varphi}$, $m$ and all nodes $m'$ are of the form $(n',0,\sharp)$ and thus all in the original model $N$. Moreover they stand in the same way in the $E$ relation. Thus we can apply the inductive hypothesis and obtain $T,n\models\until {\varphi}\psi$. As a special case, $T',(r,0,\sharp)$ satisfies ${\varphi}^\bullet$ but not $\psi^\bullet$. Recall that $T',(r,0,\sharp)$ does not satisfy the other disjuncts $AX$ of $\psi^\circ$ either and thus, $T', (r, 0, \sharp)\models {\varphi}^\bullet$ and $T' ,(r, 0, \sharp)\not\models \psi^\circ$, as desired. ($\Rightarrow$) Again we show the contrapositive. Suppose there is a model $T=(N, E, r, l)$ satisfying ${\varphi}^\bullet$ but not $\psi^\circ$ at the root $r$. Then in particular $T,r\not\models AX$. Without loss of generality, we can assume that the simulation from (the conditional tree pattern corresponding to) ${\varphi}^\bullet$ into $T$ is surjective. (Otherwise the image of ${\varphi}^\bullet$ is a subtree of $T$ and thus by monotonicity $\psi^\circ$ cannot be satisfied at the root of this subtree.) In this model, as a consequence of the fact that ${\varphi}^\bullet$ implies $\zeta$, every branch has an initial segment satisfying $\zeta$, immediately followed by a node labeled by $s$, which is followed by a segment where $\zeta$ is never satisfied because of the first disjunct of $AX$. We define a tree $T'=(N', E', r', l')$ whose set of nodes $N'$ consists of the nodes of $T$ where $\zeta$ is satisfied, i.e. $N' = \{n\in N\mid T,n\models \zeta\}$; $E'$ is simply the restriction of $E$ to $N'$ and $r'=r$. We define $l'(n)= p_i$ iff $T,n\models p_i^\bullet$, and if $T, n\not\models p_i^\bullet$ for any $i$, then we label $n$ with a fresh variable $z$. This definition of $l'$ is well-defined because the falsity of the disjunction in $AX$ ensures that each node makes at most one $p_i^\bullet$ true in $T$. Let $\theta$ be a formula in variables $\{p_1, \ldots, p_n\}$, and $n\in N$. By induction on $\theta$, we show that $$\label{eq:ctp} T,n\models\theta^\bullet\mbox{ iff } T',n\models\theta.$$ - $\theta=\top$. Then $T, n\models \top^\bullet$ $\Leftrightarrow$ (by definition of $(\cdot)^\bullet$) $T, n \models \zeta$ $\Leftrightarrow$ (by definition of $N'$) $\Leftrightarrow$ $T', n \models \top$. - $\theta=p$. Then $T, n \models p^\bullet$ $\Leftrightarrow$ (by definition of the labeling $l'$) $l'(n)= p$ $\Leftrightarrow$ $T',n \models p$. - $\theta=\neg p$. If $T, n\models (\neg p)^\bullet$, then by the falsity of $AX$, $T, n\not\models p^\bullet$, and thus $l'(n)\neq p$ and $T',n\not\models p$ and thus $T',n\models \neg p$. Conversely, $T',n\models \neg p$ $\Leftrightarrow$ $l'(n)\neq p$ $\Leftrightarrow$ $T, n\not\models p^\bullet$. But $T,n\models \zeta$ and thus either $p^\bullet$ or $(\neg p)^\bullet$ must hold at $T,n$. Thus $T, n\models (\neg p)^\bullet$. - $\theta={\varphi}_1\land {\varphi}_2$. Then $T, n\models ({\varphi}_1\land {\varphi}_2)^\bullet$ $\Leftrightarrow$ $T, n \models {\varphi}_1^\bullet$ and $T, n \models {\varphi}_2^\bullet$ $\Leftrightarrow$ ($n\in N'$ because ${\varphi}^\bullet$ implies $\zeta$ and thus by the inductive hypothesis) $T', n\models {\varphi}_1$ and $T', n\models {\varphi}_2$ $\Leftrightarrow$ $T', n\models {\varphi}_1\land {\varphi}_2$. - $\theta=\child {\varphi}$. Then $T,n \models (\child {\varphi})^\bullet$ $\Leftrightarrow$ $T, n\models \zeta$ and $T, n \models \child {\varphi}^\bullet$ $\Leftrightarrow$ there exists $n'\in N$ such that $nEn'$ and $T, n'\models {\varphi}^\bullet$ $\Leftrightarrow$ (by the fact $T, n'\models \zeta$ and, thus, $n'\in N'$ and the inductive hypothesis) there exists $n'\in N'$ such that $nE'n'$ and $T', n'\models {\varphi}$ $\Leftrightarrow$ $T', n\models \child {\varphi}$, as desired. - $\theta=\until\theta\tau$. Assume $T,n\models(\until\theta\tau)^\bullet$. Then, by definition of $(\cdot)^\bullet$, $T,n\models \zeta\land\until{\theta^\bullet}{\tau^\bullet}$. That means there exists $n'$ in $T$ with $nE^+n'$ such that $T,n'\models\tau^\bullet$ and for all $n''$ with $nE^+n''E^{+}n'$ it holds $T,n''\models \theta^\bullet$. By definition of $(\cdot)^\bullet$, the nodes $n,n'$ and all the nodes between $n$ and $n'$ satisfy $\zeta$ and thus belong to $T'$. By inductive hypothesis, $T',n'\models \tau$ and $T',n''\models \theta$ for all $nE^{+}n''E^{+}n'$, which means $T',n\models \until\theta\tau$ holds. Conversely, assume $T',n\models \until\theta\tau$. By definition of $T'$, we have that $T,n\models \zeta$, which is the first conjunct of $({\until\theta\tau})^\bullet$. The second conjunct follows by the inductive hypothesis. As $T$ is a counterexample for ${\varphi}^\bullet\subseteq\psi^\circ$, implies that $T'$ is a counterexample of ${\varphi}\subseteq \psi$, as desired. The same argument applies for multi-labeled trees. The only change is to remove the last disjunction from $AX$. Lower bounds ------------ In this section we show that the containment for $\TP^{\neg}$ is $\PSPACE$ hard. This lower bound will carry over to the containment problem for $\CTP$. 1. The containment problem for $\CTP$ is $\PSPACE$-hard. 2. The containment problem for $\TP^{\neg}$ is $\PSPACE$-hard, 3. Both results also hold for multi-labeled trees. \(i) follows from (ii) by Proposition \[prop:neg\]. (iii) follows from (i) and (ii) by Proposition \[prop:standard\]. For proving (ii), we reduce the corridor tiling problem, which is known to be hard for $\PSPACE$ [@chle:domi86; @emde:conv96], to the containment problem for $\TP^{\neg}$. We use the construction from the $\PSPACE$-hardness proof for the containment problem of $\TP$ with disjunction over a finite alphabet in [@neve:comp06]. The corridor tiling problem is formalized as follows. Let ${\sf Til}=(D,H,V, \bar{b}, \bar{t}, n)$ be a tiling system, where $D=\{d_1, \ldots,$ $d_m\}$ is a finite set of tiles, $H, V\subseteq D^2$ are horizontal and vertical constraints, $n$ is a natural number given in unary notation, $\bar{b}$ and $\bar{t}$ are tuples over $D$ of length $n$. Intuitively, given a corridor of width $n$, the goal is to construct a tiling of the corridor using the tiles from $D$ so that the horizontal and vertical constraints are satisfied and the bottom and top rows are tiled by $\bar{b}$ and $\bar{t}$, respectively. We say that a tiling satisfies the horizontal constraints $H$(respectively the vertical constraints $V$) if for every $1\leq i\leq n, j\in \nat$ it holds that if $d_1$ and $d_2$ are the tiles on the positions $(i , j)$ and $(i+1, j)$ of the corridor (respectively $(i,j)$ and $(i, j+1)$), then $(d_1, d_2)\in H$ ($(d_1, d_2)\in V$ ). Now we construct in $\PTIME$ in the length of [Til]{}, two $\TP^{\neg}$ expressions $\varphi$ and $\psi$ such that the following holds: $$\label{eq:tile} \varphi \not\subseteq \psi\ \mbox{iff there exists a tiling for}\ {\sf Til}.$$ By Proposition \[prop:mik\], we may without loss of generality construct $\psi$ as a disjunction of $\TP^{\neg}$ expressions. To this purpose, we use the string representation of a tiling. Each row of the considered tiling is represented by the tiles it consists of. If the tiling of a corridor of width $n$ has $k$ rows, it is represented by its rows separated by the special symbol $\sharp$. The top row is followed by the symbol $\$$. Thus, a tiling is a word of the form $u_1 \sharp u_2 \sharp \cdots \sharp u_k \$$, where each $u_i$ is the word of length $n$ corresponding to the $i$-th row in the tiling. In particular $u_1=\bar{b}$ and $u_k=\bar{t}$. For the sake of readability, for expression $r$, $r^i$ denotes the path formula $?r;\downarrow; ?r; \ldots; \downarrow; ?r$ with $i$ occurrences of $r$. Let $\varphi$ be $$\tup{?b_1;\downarrow; ?b_2 ; \ldots ; \downarrow; ?b_n; \downarrow ; ?\sharp;\ \downarrow^+\ ; ?t_1,\downarrow; \ldots \downarrow ; ?t_n;\downarrow}\$.$$ Intuitively, this expression enforces a tiling to start with a path starting with $\bar{b}$ and finishing with $\bar{t}$ and the final symbol $\$$. Now the formula $\psi$ defines all incorrect tilings and additional constraints. It is the disjunction of the following $\TP^{\neg}$ formulas. 1. $\tup{\downarrow^+\ ; ?(\neg d_1\wedge \ldots \wedge \neg d_m\wedge \neg \sharp);\ \downarrow^+}\$$. There is a position that is labeled neither by a tile nor a delimiter. 2. Incorrect length of a row. 1. $\bigvee^{n-1}_{i=0}\tup{\downarrow^+\ ;?\sharp; \top^i;\ \downarrow}\sharp$, a row is too short; 2. $\tup{\downarrow^+\ ;(\neg \sharp)^{n+1}} \top$, a row is too long; 3. $\bigvee_{(d_1,d_2)\not\in H}\tup{\downarrow^+; ?d_1; \downarrow ; ?d_2} \top$, some horizontal constraints are violated; 4. $\bigvee_{(d_1,d_2)\not \in V}\tup{\downarrow^+; ?d_1; \downarrow; \top^n; \downarrow; ?d_2}\top$, some vertical constraints are violated. Note that negated labels are used in (0) and (1b). Also note that the size of $\varphi$ and $\psi$ is bounded by a polynomial in the size of [Til]{}. We now show . $(\Leftarrow)$. Assume that there exists a tiling of the corridor. Let $s$ be the string representation of it. Then, $s=u_1\sharp u_2 \sharp \dots \sharp u_k \$$, where $|u_i|=n$, $u_i \in D^n$, $u_1=\bar b$ and $u_k= \bar t$. Moreover, on the one hand if $x\cdot y$, is an infix of some $u_i$, then $(x,y) \in H$, and on the other hand for every infix $x\cdot u' \cdot y$ of length $n+1$ of $u_i \sharp \cdot u_{i+1}$, it holds that $(x,y) \in V$. Let $T_s$ be the corresponding tree, i.e. a single path of $|s|$ nodes $\{v_1, \ldots , v_{|s|}\}$ where the labeling is set in accordance with $s$, i.e. $l(v_i)=s_i$. Clearly, $T_s$ is a model of $\varphi$ and not of $\psi$. $(\Rightarrow)$. Let $T$ be a tree such that $T, r\models \varphi$ and $T, r\not\models \psi$. Since $T, r\models \varphi$, there must exist a path $r=v_1,\ldots, v_m$ in $T$ which starts with $\bar{b}$ and finishes with $\bar{t}\$$. Moreover, either $\sharp$ or a symbol from $D$ is the label of every node $v_i, 1\leq i< m$, according to (0). We define a tiling function $g: \{0, \ldots, n-1\}\times \nat\to D$ assigning a tile to every position in the corridor as follows: $g(i,j)=l(v_{(n+1)\times j+i+1}), 1\leq i\leq n$, where $l$ is the labeling function of $T$. Indeed, this function is well defined, as (1) ensures the correct counting. By formulas (2) and (3) the tiling defined by $g$ satisfies the horizontal and vertical constraints. The difference between $\CTP$ patterns and $\TP$ patterns is that $\CTP$ descendent edges can be labeled by patterns. One might ask whether a bound on the degree of such a labeling nesting can lead to a lower complexity of the containment problem. Define the until nesting depth $un: \CTP \to \mathbb N$ as follows. - $un(p)=un(\top)=0$, - $un(\tup{\downarrow}{\varphi})=un(\until{\downarrow}{{\varphi}})=1$, - $un({\varphi}_1\land {\varphi}_2)=\max\{un({\varphi}_1),un({\varphi}_2)\}$, - $un(\until{{\varphi}}{\psi})=un({\varphi})+1$. Unfortunately, a close examination of the encoding in the lower bound proof and the encodings in Propositions \[prop:mik\] and \[prop:neg\] gives a negative answer to the question. Thus we obtain The $\CTP$ containment problem for formulas of until nesting depth one is $\PSPACE$-hard. Upper bounds {#sec:ctp} ------------ In this section, we show a matching $\PSPACE$ upper bound for $\CTP$ containment. The complexity of $\CTP^{\neg,\vee}$ containment follows from a translation into existential $\CTL$ ($\ECTL$), whose containment problem is known to be $\PSPACE$-complete [@kupf:auto00]. The only small technical issue is that $\ECTL$ formulas are interpreted over *infinite* finitely branching trees. We solve that by relativizing formulas with a new propositional variable $s$, whose interpretation will provide the desired finite tree. \[thm:ctp-up\] The containment problem for $\CXP^{\neg,\vee}$ is decidable in $\PSPACE$. This also holds for multi-labeled trees. We prove the upper bound for the case of multi-labeled trees. The upper bound for single-labeled trees then follows by Proposition \[prop:standard\]. We recall from [@kupf:auto00] the definition of $\ECTL$. Let $\pvar$ be a set of propositional variables. $\ECTL$ node formulas ${\varphi}$ and path formulas $\alpha$ are defined by: $$\begin{aligned} {\varphi}&::= \top \mid \bot \mid p \mid \lnot p \mid {\varphi}\land {\varphi}\mid {\varphi}\lor {\varphi}\mid \exists \alpha \\ \alpha &::= {\varphi}\mid \alpha \land \alpha \mid \alpha \lor \alpha \mid X \alpha \mid (\alpha U \alpha) \mid (\alpha \hat{U} \alpha)\end{aligned}$$ where $p \in \pvar$, and asking that in node formulas, the operators $X$(“next"), $U$(“until") and $\hat{U}$(“dual of until") are always immediately preceded by $\exists$. The semantics for $\ECTL$ is given by *infinite finitely branching trees*. $\ECTL$ path formulas are interpreted at a (possibly infinite) paths of the tree and $\ECTL$ node formulas are interpreted at nodes of the tree. Our reduction will translate $\CTP^{\neg , \vee}$ formulas to $\ECTL$ node formulas, with occurrences of formulas of the form $\exists X {\varphi}$ and $\exists \psi U {\varphi}$. Semantics of such formulas is defined as follows, given an infinite tree $T=(N, E, r, \rho)$ and a node $n\in N$. $$\begin{aligned} T, n\models \exists X{\varphi}\ & \text{iff} \ \text{there exists a node }m\in N \ \text{such that}\ nEm\ \text{and}\ T, m\models {\varphi}, \\ T, n\models \exists \psi U {\varphi}\ & \text{iff}\ \text{there exists a node}\ m\in N \ \text{such that}\ nE^*m\ \text{and}\ T, m\models {\varphi},\\ & \text{and} \ \text{for all}\ n'\in N \ \text{with}\ nE^*n'E^+m \ \text{it holds}\ T, n'\models \psi. \end{aligned}$$ The $\CTP$ node formula $\until{{\varphi}}{\psi}$ corresponds to the strict until operator, which is expressible in $\ECTL$ as $\exists X\exists \psi U {\varphi}$. We denote by $\subseteq_\infty$ the containment relation for $\ECTL$ node formulas over finitely branching trees where each branch is infinite. Let ${\varphi},\psi$ be in $\CXP^{\neg,\vee}$ and let $s$ be a new propositional variable not in ${\varphi},\psi$. The translation $(\cdot)^s$ from $\CXP^{\neg,\vee}$ to $\ECTL$ node formulas relativizes every subformula with $s$ and adjusts the syntax. $$\begin{aligned} (\top)^s &=& s\land \top \\ (p)^s& = & s\land p \\ (\neg p)^s &=&s\land \neg p \\ ({\varphi}_1\land {\varphi}_2)^s & = & {\varphi}^s_1 \land {\varphi}^s_2 \\ ({\varphi}_1\lor {\varphi}_2)^s &=& {\varphi}^s_1\lor {\varphi}^s_2 \\ (\tup{\downarrow}{\varphi})^s &=& s \land \exists X {\varphi}^s \\ (\until{\psi}{{\varphi}})^s &=& s \land \exists X \exists \psi^sU{\varphi}^s\end{aligned}$$ We claim that ${\varphi}\subseteq \psi$ iff ${\varphi}^s \subseteq_\infty \psi^s$. The proof is by contraposition. First assume that for some finite tree $T=(N, E, r, \rho)$ and node $n\in N$ it holds $T,n\models{\varphi}$ and $T,n\not\models\psi$. We then construct an infinite tree which is a counterexample. Let $s$ be a propositional variable not occurring in ${\varphi}$ or $\psi$. Let $T^s_\infty$ be the infinite tree obtained from $T$ by adding to each leaf in $T$ an infinite path. The labeling is changed as follows: all new nodes have the empty label set and we add the label $s$ to the label set of all old nodes. Formally, $T^s_\infty=(N_\infty, E_\infty, r, \rho_\infty)$, where $N_\infty=N\cup \{n^m_1,n^m_2, \ldots \mid m \text{ is a leaf in }T\}$ is the set of nodes, $E_\infty=E\cup \{(m, n^m_1)\mid n^m_1\in N_\infty\}\cup \{(n^m_i,n^m_{i+1})\mid n^m_i, n^m_{i+1}\in N_\infty, i\geq 1\}$ is the set of edges, $r$ is the root and $\rho_\infty$ is the labeling function defined as follows. $$\rho_\infty(n) = \left\{ \begin{array}{l l} \rho(n)\cup \{s\} & \quad \text{if\ }n\in N, \\ \emptyset & \quad \text{if\ } n\in N_\infty\setminus N. \end{array} \right.$$ We show by induction on the structure of the formula that for every node $n\in N$ and every $\CTP^{\neg,\vee}$ formula $\theta$, $T,n\models \theta$ iff $T^s_\infty, n\models \theta^s$. To show this we will use the following two facts which are easy consequences of the definitions of $T^s_\infty$ and the translation $(\cdot)^s$: $$T^s_\infty, n \models s \ \Leftrightarrow \ n\in N, \eqno{(*)}$$ and $$T^s_\infty, n\models {\varphi}^s \ \Rightarrow \ T^s_\infty,n \models s. \eqno{(**)}$$ - $\theta=\top$. We have $T, n \models \top$ iff $n\in N$ iff $n\in N_\infty$ and $s\in \rho_\infty(n)$ iff $T^s_\infty, n \models s\land \top$ iff $T^s_\infty, n\models (\top)^s$. - $\theta=p, p\in \Sigma$. We have $T, n\models p$ iff $p\in \rho(n)$ iff (since $n\in N$) $\{p, s\}\subseteq \rho_\infty(n)$ iff $T^s_\infty, n\models s\land p$ iff $T^s_\infty, n\models (p)^s$. - $\theta=\neg p, p\in \Sigma$. We have $T, n\models \neg p$ iff $p\not \in \rho(n)$ iff (since $n\in N$ and, thus, $s\in \rho_\infty(n)$) $p\not \in \rho_\infty(n)$ and $s\in \rho_\infty(n)$ iff $T^s_\infty, n\models s\land \neg p$ iff $T^s_\infty, n\models (\neg p)^s$. - $\theta={\varphi}_1\land {\varphi}_2$. We have $T, n\models {\varphi}_1\land {\varphi}_2$ iff $T,n\models {\varphi}_1$ and $T, n\models {\varphi}_2$ iff (by the induction hypothesis, (\*) and (\*\*)) $T^s_\infty, n\models {\varphi}^s_1$ and $T^s_\infty, n\models {\varphi}^s_2$ iff $T^s_\infty, n\models {\varphi}^s_1\land {\varphi}^s_2$ iff $T^s_\infty, n\models {\varphi}^s$. - $\theta={\varphi}_1\lor {\varphi}_2$. By the same argument as for ${\varphi}_1\land {\varphi}_2$. - $\theta=\tup{\downarrow}{\varphi}$. We have $T, n\models \tup{\downarrow} {\varphi}$ iff there exists $m\in N$ such that $nEm$ and $T, m\models {\varphi}$ iff (by the induction hypothesis, (\*), (\*\*) and the fact that $n\in N$) $T^s_\infty, n\models s$ and there exists $m\in N_\infty$ such that $nE_\infty m$ and $T^s_\infty, m\models {\varphi}^s$ iff $T^s_\infty, n\models s\land \exists X {\varphi}^s$ iff $T^s_\infty \models (\tup{\downarrow}{\varphi})^s$. - $\theta=\until{\psi}{{\varphi}}$. We have $T, n \models \until{\psi}{{\varphi}}$ iff there exists $m\in N$ such that $T, m\models {\varphi}$ and for all $n'\in N$ with $nE^+ n' E^+ m$ it holds $T, n'\models \psi$. Then by the induction hypothesis, (\*) and (\*\*), the latter is equivalent to existence of $m\in N_\infty$ such that $T^s_\infty,m \models {\varphi}^s$ and for all $n'\in N_\infty$ with $nE^+_\infty n' E^+_\infty m$ it holds $T^s_\infty \models \psi^s$. The latter implication holds because for every node $n'$ with $nE^+_\infty n' E^+_\infty m$ it holds $n'\in N$ and $nE^+n'E^+m$. Equivalently, there exists $n_1\in N_\infty$ and $m\in N_\infty$ such that $nE_\infty n_1E^*_\infty m$, $T^s_\infty, m \models {\varphi}^s$ and for all $n'$ with $n_1E^*_\infty n' E^+_\infty m$ it holds $T^s_\infty, n'\models \psi^s$. This is equivalent to $T^s_\infty, n \models s\land \exists X\exists \psi^s U {\varphi}^s$ as desired. Thus, we obtain both $T^s_\infty, n \models {\varphi}^s$ and $T^s_\infty,n \not\models \psi^s$. For the other direction, let $T_\infty=(N_\infty, E_\infty, r, \rho_\infty)$ be some infinite tree over $\Sigma$ such that $T^\infty,r\models {\varphi}^s$ and $T^\infty,r\not\models \psi^s$. We then remove from the model all nodes without $s$ and their descendants. Formally, $T=(N, E, r, \rho )$, where $N=N_\infty \setminus \{n\mid \exists n' \in N_\infty. s\not\in \rho_\infty(n')\land n'E^*_\infty n\}$ is the set of nodes, $E=E_\infty|_{N\times N}$ is the set of edges, $r$ is the root, and $\rho(n)=\rho_\infty(n)$ is the labeling function. By induction we can show that for every $\CTP^{\neg, \vee}$ formula $\theta$ and node $n\in N$, it holds $T_\infty, n\models \theta^s$ if and only if $T, n\models \theta^s$. The two non-trivial cases are the following. - $\theta = \tup{\downarrow}{{\varphi}}$. We have $T_\infty ,n \models s\land \exists X {\varphi}^s$ iff $T_\infty, n\models s$ and there exists $m\in N_\infty$ such that $nE_\infty m$ and $T_\infty, m\models {\varphi}^s$ iff $T, n \models s$ and there exists $m\in N$ such that $nEm$ and $T, m\models {\varphi}^s$. The direction from right to left is obvious, while the direction from left to right because of the following. Since $T_\infty, m\models {\varphi}^s$, it follows that $T_\infty, m \models s$, i.e. $s\in \rho_\infty(m)$. Then since $n\in N$, we have that $m\in N$ too by the definition. The latter also implies that $nE m$ holds in $T$ and thus that $T, n\models s\land \exists X {\varphi}^s$. - $\theta=\until{\psi}{{\varphi}}$. We have $T_\infty , n \models s\land \exists X\exists \psi^s U{\varphi}^s$ iff $T_\infty, n\models s$ and there exists $m\in N_\infty$ such that $T_\infty , m\models {\varphi}^s$, $nE^+_\infty m$ and for all $n'\in N_\infty$ with $nE^+_\infty n'E^+_\infty m$ it holds $T_\infty, n' \models \psi^s$. This is equivalent to $T, n \models s$ and there exists $m\in N$ such that $T, m\models {\varphi}^s$ and for all $n'$ with $nE^{+}n' E^{+} m$ it holds $T, m \models \psi^s$, which means $T, n\models s\land \exists X\exists \psi^s U{\varphi}^s$ as desired. The direction from left to write is trivial since $T$ is a substructure of $T_\infty$. The direction from right to left follows from the fact that $m\in N$ and all the nodes $n'$ such that $nE^+_\infty n' E^+_\infty m$ belong to $N$ as well and, moreover, it holds $nE^{+} n' E^{+} m$. Thus, we obtain that $T, r\models {\varphi}^s$ and $T, r\not\models \psi^s$. Because all nodes of $T$ now make $s$ true, we can discard the relativization with $s$ in the formulas. Thus obtaining $T,r \models {\varphi}$ and $T, r\not\models \psi$. The only problem is that $T$ is infinite. But using the simulations from Theorem \[prop:simc\] it is easy to see that we can turn $T$ into a finite counterexample. Thus, we have shown that ${\varphi}\subseteq \psi$ if and only if ${\varphi}^s\subseteq_\infty \psi^s$. The latter containment problem in $\ECTL$ is known to be decidable in $\PSPACE$. This concludes the proof. The containment problem for $\TP^{\neg}$ is $\PSPACE$ complete. However, restricting negation to a safe negation $p\land \neg q_1\land \ldots \land \neg q_n$ keeps the complexity of the containment problem for tree patterns in $\coNP$. Note that $p\land \neg q_1\land \ldots \land \neg q_n$ only adds expressive power over multi-labeled models, because on single labeled models $p\equiv p\land\neg q_1\land \ldots \land \neg q_n$. The proof of the following result can be found in [@marx:cont14]. Here $\TP^{\lor, \neg^s}$ is tree patterns expanded with disjunction and the safe negation construct. The containment problem for $\TP^{\lor, \neg^s}$ over multi-labeled trees is in $\coNP$. Conclusion ========== We have shown that adding conditions on the edges of tree patterns gives a boost in expressive power which comes with the price of a higher – $\PSPACE$ – complexity for the containment problem than for tree patterns. We located the source of the extra complexity in the fact that unrestricted negations of labels can be coded in Conditional Tree Patterns. Adding negations of labels to tree patterns causes an increase in complexity of the containment problem from $\coNP$ to $\PSPACE$. This paper is a first step in exploring “Regular Tree Patterns” and fragments of it. We mention some directions for future work. Miklau and Suciu in [@mikl:cont04] mention that existence of a homomorphism between tree patterns is a necessary but not sufficient condition for containment in $\TP$. Can we extend the simulations between conditional tree patterns and trees to simulations between queries, partly capturing containment as for tree patterns? What is the complexity of containment of regular tree patterns, i.e. the positive fragment of Regular XPath without disjunction and union? As the satisfiability (and thus also the containment) problem for Regular XPath is known to be $\EXPTIME$-complete, it must lie in between $\PSPACE$ and $\EXPTIME$. There are also interesting characterization questions: what is the exact $\mathbf{FO}$ fragment corresponding to $\CXP$ or $\CXP$ with disjunction and union? Ten Cate [@cate:expr06] showed that full Regular XPath (with all four axis relations) expanded with path equalities is equally expressive as binary $\mathbf{FO}^*$ (First-order logic extended with a transitive closure operator that can be applied to formulas with exactly two free variables). Is every positive forward binary $\mathbf{FO}^*$ formula expressible as a Regular Tree Pattern? Is $\CXP$ with disjunction and union equally expressive as $\mathbf{FO}$ intersected with positive $\mathbf{FO}^*$? Are unions of $\CXP$ equivalent to unions of the first order fragment of conjunctive regular path queries [@calv:cont00]? Acknowledgement =============== We would like to thank the anonymous referees for their helpful comments. Translations between $\CTP$ and $ctp$ {#app:tra} ===================================== The following constructions are similar to the tree and graph representations for tree patterns and conjunctive queries over trees [@mikl:cont04; @bjor:conj11] We define the translation function $\ct(\cdot)$ which assigns an equivalent conditional tree pattern with one output node to every $\CTP$ formula. The output node of $\ct({\varphi})$, where ${\varphi}$ is a node formula, equals the root. Let $\alpha$ and ${\varphi}$ be path and node $\CTP$ formulas. We define $\ct(\alpha) $ and $\ct({\varphi})$ by mutual induction on the complexity of the path and node formulas. We take $\ct(\alpha)$ or $\ct({\varphi})$ to be the tree $ (N, E , r, o, \rho_N, \rho_E)$, where the components are defined according to the cases. - $\alpha = \downarrow$. Then $N$ consists of two nodes $r$ and $n$. The edge relation $E$ is defined as $ \{\tup{r, n}\}$. Moreover, $o := n$, $\rho_N(v) = \emptyset$ for $v\in \{r, n\}$ and $\rho_E(\tup{r,n}) = \downarrow$, - $\alpha = ? {\varphi}$. Then $\ct(\alpha) := \ct({\varphi})$, - $\alpha = \alpha_1 ; \alpha_2$. Let $\ct(\alpha_1)$ and $\ct(\alpha_2)$ be the conditional tree patterns for $\alpha_1$ and $\alpha_2$. Then $c(\alpha)$ is the tree obtained as follows. We fuse the root of $\ct(\alpha_2)$ with the output node of $\ct(\alpha_1)$ and declare the output node of $\ct(\alpha_2)$ as the output node of $\ct(\alpha)$. The label of the fusion node is the union of the labels of the output node of $\ct(\alpha_1)$ and the root of $\ct(\alpha_2)$. Labels of other nodes and the edges remain the same as in $\ct(\alpha_1)$ and $\ct(\alpha_2)$. - $\alpha = \untilpath$. Then $N$ consist of two nodes $r$ and $n$. The edge relation $E$ is defined as $ \{\tup{r, n}\}$. Moreover, $o := n$, $\rho_N(v) = \emptyset$ for $v\in \{r, n\}$ and $\rho_E(\tup{r,n}) = \ct({\varphi})$. For node formulas, the output node of the translation result is always defined as the root $r$. - ${\varphi}= p$. Then $N$ consists of a single node $r$, $E$ is empty, the labeling $\rho_N(r)=\{p\}$. - ${\varphi}= \top$. Similar to the previous case, with the exception that $\rho_N(r) = \emptyset$ - ${\varphi}= {\varphi}_1 \wedge {\varphi}_2$. Then $\ct({\varphi}) := \ct({\varphi}_1)\oplus \ct({\varphi}_2)$, i.e. the fusion of the conditional tree patterns $\ct({\varphi}_1)$ and $ \ct({\varphi}_2)$. - ${\varphi}= \tup{\alpha} {\varphi}_1$. Let $\ct(\alpha)$ and $\ct({\varphi}_1)$ be the corresponding conditional tree patterns. Then $\ct({\varphi})$ is obtained by fusing the output node of $\ct(\alpha)$ with the root of $\ct({\varphi})$. The labeling of the fusion node is defined as the union of the labels of the root of $\ct({\varphi})$ and the output node of $\ct(\alpha)$. Labels of the other nodes and edges remain the same as in $\ct(\alpha)$ and $\ct({\varphi})$. Translation $f(\cdot)$ works the other way around. Let $t= (N, E , r, o, \rho_N, \rho_E)$ be a conditional tree pattern with one output node. We define $f(t)$ by induction on the nesting depth and the depth of the tree. We first define a mapping ${\varphi}$ from nodes $v\in N$ to $\CTP$ node formulas. The mapping is defined inductively starting from the leaves as follows. Here, for a finite set $S= \{p_1, \ldots, p_n\}$, by $\wedge S$ we denote the finite conjunction $p_1\wedge \ldots \wedge p_n$. We take $\wedge \emptyset$ to be $\top$. For a conditional tree pattern $t$, by $r_t$ we denote the root of $t$. For $v$ a node in pattern $t$: $$\begin{aligned} {\varphi}(v) &=\wedge \rho_N(v)\ \wedge \\ & \bigwedge_{\tup{v, v'}\in E ,\ \rho_E(v, v')=t'} \tup{? {\varphi}(r_{t'})}{\varphi}(v')\ \wedge \\ & \bigwedge_{\tup{v, v'} \in E , \ \rho_E(v, v') = \downarrow} \child{{\varphi}(v')}\end{aligned}$$ Now let $r=v_1, \ldots, v_n=o$ be the path from the root $r$ to the output node $o$ in $t$. Let the expression $d_i, 1\leq i\leq n-1$ be defined by: $d_i = \downarrow$ if $\rho_E(v_i, v_{i+1}) =\downarrow$ and $d_i = \upath{{\varphi}(r_{t'})}$ if $\rho_E(v_i, v_{i+1}) = t'$. Then the result of the translation $f(t)$ of the conditional tree pattern $t$ is the $\CTP$ path formula: $$?{\varphi}(v_1); d_1 ; ?{\varphi}(v_2); d_2; \ldots; ?{\varphi}(v_{n-1}); d_{n-1} ;? {\varphi}(v_n).$$ For the next proposition we need the definition of equivalence between conditional tree patterns. Let $t$ be a conditional tree pattern with output nodes $\bar{o}$,$ |\bar{o}|$=$k$, and $T$ a tree. Then the *answer set* of $t$ over $T$ is the set $Out(t, T)=\{\tup{g(o_1), \ldots, g(o_k)}\mid g \text{ is a simulation of } t \text{ in } T\}$. We say that two conditional tree patterns $t_1$ and $t_2$ are *equivalent*, denoted as $t_1\simeq t_2$, if $Out(t_1, T) = Out(t_2, T)$ for every tree $T$. Let ${\varphi}$ and $\alpha$ be $\CTP$ node and path formulas, $t$ a conditional tree pattern. Then it holds that 1. $f(\ct({\varphi})) \equiv {\varphi}$ and $f(\ct(\alpha)) \equiv \alpha$, 2. $\ct(f(t)) \simeq t$. The proof of (i) is by mutual induction on $\alpha$ and ${\varphi}$, and the proof of (ii) is by induction on nesting depth and depth of $t$. [^1]: This research was supported by the Netherlands Organization for Scientific Research (NWO) under project number 612.001.012 (DEX). | Mid | [
0.6366197183098591,
28.25,
16.125
] |
HulloMail launches voicemails you can read HulloMail Scribe voicemail to text service now on Android and iPhone Shares If you have ever groaned or moaned about the time it takes you to dial-in and pick up your voicemails, then a new app service from HulloMail is just what you have been waiting for. The Hullomail voicemail-to-text service allows you to read your voicemails quickly via the app, so you only have to answer the urgent ones, and you no longer have to dial-in, press endless buttons and then listen to long-winded messages. The new version of the HulloMail's SmartVoicemail app has been enhanced, with the addition of a new voice-to-text transcription technology, Scribe, that gives users the option to have a 10-second 'snapshot' of voicemails transcribed into text format within the app. The voicemail transcriptions are fully automated and use proven voice-to-text transcription technology that has been integrated in partnership with Nuance – they're the people behind Dragon Dictation on the iPhone and Dragon Naturally Speaking . The new Get The Gist product is available for HulloMail's Android and iPhone subscribers for a monthly subscription of £1.49 in the UK, $1.99 in the USA and €1.59 in Ireland. For users of HulloMail's free Android app, the monthly subscription is £1.99 in the UK, $2.99 in the USA and €2.39 in Ireland. This will be extended to the free HulloMail iPhone application in the next client update. The subscriptions entitle a user to have up to 50 of their voicemails converted to text. | Mid | [
0.6438356164383561,
35.25,
19.5
] |
The regulation of cell-mediated immunity (CMI) is being investigated in the context of chronic infections, in general, and leprosy, in particular. A number of experimental mouse models are being examined. Some, such as vaccination with living BCG or irradiation-killed M. leprae induce a highly stable state of CMI. Conversely, subcutaneous infection with M. lepraemurium (MLM) engenders a very unstable state of CMI which is easily suppressed by systemic injection of either live, intact dead, or disrupted MLM. Similar suppression occurs spontaneously during the course of normally progressive infection. Stable states of CMI will also be specifically suppressed by systemic injections of homologous antigen. Unstable CMI will be averted or reversed by treatment with immunopotentiating agents. Preliminary evidence suggests that the suppression of CMI is mediated by sub-classes of T-lymphocytes, and these cells will be characterized by use of Lyt and I-J antisera and other cell separation techniques. These procedures will also be used to determine the identity of the cells which mediate delayed hypersensitivity, antigen-specific lymphocyte transformation, lymphokine production and protective immunity. | Mid | [
0.647342995169082,
33.5,
18.25
] |
Q: Is this formula for Standard Error just for a sampling distribution that's normal? This basic formula of finding the Standard Error (Or Standard Deviation) of a Sampling Distribution, S.E= s/sqrt(n) Where s= std deviation of sample n=sample size Is this only for the Sampling Distributions of "Means" alone and of those that are Normally Distributed? Or does it apply to other point estimates, or other distributions? A: It is the standard deviation of the sample mean for any distribution, not just the normal. This follows from the theory of cumulants. Also, it is exact for any $n$, not just asymptotic. It is not the standard deviation of any sample moment or other sample aggregate quantity, just of the mean. | High | [
0.670360110803324,
30.25,
14.875
] |
Q: UART pins to unpowered MCU? I'm using a USB-UART IC with an MCU, which is powered from a battery. The USB-UART IC is powered from USB connector, not from the battery, so that I don't need to open up a console every time when the switch goes off and on. simulate this circuit – Schematic created using CircuitLab edit: I didn't draw it on the figure but the IC has an internal 3.3V regulator and every VDD is on the 3.3V level when the switch is on. Now I'm worried about when the USB is plugged and the switch is still off. The MCU document says that every input pin's maximum rating is VDD + 0.3, which would be 0.3 V when the MCU is not powered. If the TX/RX pair on the USB-UART side goes high, will it destroy the pins on the MCU side? If so, What do I need between the TX/RX pairs? A: It depends on the MCU, but in most cases it'll power up the MCU, and possibly the rest of the board through the MCU. The MCU will try to run, and do odd things. Your board will do odder things. If your board draws enough current, it'll damage that pin on the MCU. You need to arrange for the UART signal to stay at 0V when the MCU is off. If the UART chip (or UART) that you are using doesn't have an enable pin (the USB UART chips that I've worked with can be configured for exactly the case you're describing), then AND the UART outputs with the microprocessor's VCC. A: The solution might be as simple as an NMOS between the TX pin of the USB-UART (drain) and the RX pin of the MCU (source), with the gate connected to the MCU VDD. This will mean that the USB-UART will only be able to drive up to 3.3 V minus the threshold voltage of the transistor. You'll have to check whether this is still enough to satisfy the V_IH of your MCU. I'm skeptical of solutions that involve a logic gate powered (indirectly) by VBUS. As long as the logic gate drives the MCU input high, the MCU VCC may not drop far enough to switch off the TX driver. With the NMOS solution, the TX pin can only be driven to a lower voltage than VCC, making this sort of feedback impossible. By the way, you should also consider the reverse direction: If the MCU is powered, but the USB-UART is not, you'll want to make sure that you don't accidentally provide some power to the USB-UART, draining your battery faster than you expect. A software-only solution should suffice for this. | Mid | [
0.607317073170731,
31.125,
20.125
] |
493 F.Supp. 938 (1980) Keith D. CIUMMEI v. Ronald W. AMARAL et al. Civ. A. No. 79-46-T. United States District Court, D. Massachusetts. July 9, 1980. *939 Robert W. Hagopian, John B. Miller, c/o Orion Research Inc., Cambridge, Mass., for plaintiff. Barbara A. H. Smith, Asst. Atty. Gen., Chief, Criminal App. Section, Boston, Mass., for defendants. OPINION TAURO, District Judge. Petitioner seeks habeas corpus relief from his 1975 bench trial conviction in the Massachusetts Superior Court for unarmed robbery. Ciummei filed, but did not pursue, an appeal from that conviction. A subsequent application for sentence review was filed but later withdrawn. In 1977, he filed a petition for a writ of error challenging the voluntariness of his jury waiver. That matter was referred to a special master. After examining the trial record, as well as holding an evidentiary hearing, the master recommended affirmance of the conviction. That recommendation was accepted by a single justice of the Supreme Judicial Court and later affirmed by the full bench. Ciummei v. Commonwealth, ___ Mass. ___, 392 N.E.2d 1186 (1979). Here petitioner advances three theories in support of his request for relief. First, he contends that the trial judge's failure to explain his right to a jury trial, and the consequences of a waiver, was error of constitutional magnitude. Second, he argues that his waiver in this case was in fact not voluntarily and knowingly made. Third, he attacks on equal protection grounds the Supreme Judicial Court's refusal to apply the in-court colloquy requirement announced in Ciummei, supra, to Ciummei himself. I. The trial judge's responsibility is to assure that a defendant's proffer of a jury waiver is both knowing and voluntary. See Adams v. United States ex rel. McCann, 317 U.S. 269, 275, 63 S.Ct. 236, 240, 87 L.Ed. 268 (1942). The immediate question is whether in seeking to meet that responsibility, the constitution requires the judge to conduct a contemporaneous colloquy with the defendant. A few jurisdictions have found a constitutional mandate for such a colloquy, but most have not.[1] Essentially, the majority hold that, for due process purposes, knowledge and voluntariness may be demonstrated by reference to the entire record including evidence extrinsic to the waiver hearing. A contemporaneous colloquy would be relevant evidence of knowledge and voluntariness, but not constitutionally required under the majority view. In Ciummei, supra, the Supreme Judicial Court clearly put Massachusetts in the majority column, finding no constitutional right to a contemporaneous hearing. This court concurs with Justice Kaplan's thoughtful analysis of the issue and holds that there is no due process requirement for a contemporaneous colloquy in order to validate a proffered jury waiver.[2] *940 In those cases where a contemporaneous colloquy or hearing has been held to be constitutionally required, the underlying issue was whether there would be any trial of the merits. See Boykin v. Alabama, 395 U.S. 238, 89 S.Ct. 1709, 23 L.Ed.2d 274 (1969) (guilty plea); Commonwealth v. Hill, 375 Mass. 50, 375 N.E.2d 1168 (1978) (mental competency). The issue here is substantively different. That petitioner would be tried was never at issue. The only question was whether trial would be by bench or jury. In Massachusetts, that choice is left to the defendant, subject to court approval.[3] The issue of jury waiver is primarily one of trial strategy. Waiver has no adverse effect on a defendant's rights to confront witnesses, avoid self-incrimination, and be judged by an impartial factfinder. See, Ciummei, supra, at 1189; Boykin, 395 U.S. supra, at 243, 89 S.Ct. at 1712. Having these distinguishing factors in mind, this court agrees with Justice Kaplan's conclusion that ". . . it is not unreasonable for courts to refrain, in the case of the jury right, from constitutionalizing a particular means of demonstrating the legality of the waiver." Ciummei, supra, at 1189. Petitioner's fallback position is that the majority view concerning due process standards at a waiver hearing is inapposite when, as here, the trial judge is aware that the defendant has limited mental capacity. In that situation, petitioner argues, due process requires a contemporaneous colloquy before a jury may be waived. He analogizes the waiver under such a circumstance to the procedure mandated for testing the competency of a defendant to stand trial. See Pate v. Robinson, 383 U.S. 375, 86 S.Ct. 836, 15 L.Ed.2d 815 (1966) (mandating evidentiary hearing where judge has reason to doubt defendant's competency to stand trial); Hill, supra (same). Petitioner correctly identifies the mental competency of a defendant as a relevant factor for the trial court to weigh in assessing the validity of a proffered jury waiver. His argument, however, that the added issue of competence is a matter of constitutional significance in a jury waiver setting is unpersuasive. Certainly, a trial judge must be satisfied that a defendant is not "rendered incapable of rational judgment." Ciummei, supra, at 1190. But, the issue of competence can be resolved on the basis of evidence independent of a contemporaneous colloquy.[4] Given a question of competency, a defendant does not even get to the point of proffering a jury waiver until the trial court has determined that the defendant is fit to stand trial. Pate, supra. Assuming a determination of competence, a defendant's subsequent decision to proffer a jury waiver is best characterized as a trial tactic. Against this background, the court concludes that constitutional requirements are met when the record as a whole supports a judge's determination that a competent defendant has waived a jury trial knowingly and voluntarily. The constitution does not mandate a particular ritual as a predicate to such a determination. II. Petitioner next asserts the factual argument that his waiver was neither knowing nor voluntary. Under 28 U.S.C. § 2254(d), facts found by the state court and the special master are conclusive, absent *941 circumstances not present in this case.[5] The duty of a federal court on a petition for habeas corpus is to make an independent determination of the legal effect of such facts. Brewer v. Williams, 430 U.S. 387, 397 n. 4, 403-04, 97 S.Ct. 1232, 1239, 1242, 51 L.Ed.2d 424 (1977) (waiver of counsel right is a question of federal law applied to state court findings of fact). The relevant facts as established in the state proceedings have been summarized in Ciummei, supra, at 1187-88 and 1190-92.[6] Among the indicia of a knowing and voluntary waiver therein recited are: (1) a jury waiver form signed by petitioner and submitted to the trial court at the beginning of trial, see U. S. ex rel. Gentry v. Circuit Court of Cook County, 586 F.2d 1142, 1145 (7th Cir. 1978); (2) testimony by petitioner's trial counsel that he evidenced no lack of understanding about his right to a jury trial in other criminal proceedings; (3) evidence that both the trial judge and petitioner's counsel had handled his earlier criminal prosecutions and were aware of his history of mental problems; (4) expert medical testimony supporting a finding of competency to stand trial; (5) counsel's explanation to petitioner as to the advantages of proceeding jury-waived in the face of the Commonwealth's refusal to dismiss the charge of armed robbery;[7] (6) petitioner's previous court experiences (though none of these was a jury trial); and (7) petitioner's rational and articulate trial testimony. Petitioner assails reliance on any of these indicia on several grounds. First, he argues that his counsel failed to explain the nature of the jury its composition, the requirement of unanimity, the defendant's participation in empanelling, etc. Second, petitioner recites the master's statement that he "was in some degree confused as to the function of the jury." Third, he points to his record of emotional disturbance and mental retardation as further aggravating the pernicious effects of the failures to inform him completely about the right he waived. Finally, petitioner attacks his trial counsel's failure to make a recommendation as to which course he should follow. This court is not satisfied that these arguments outweigh the evidence supporting the state court's findings of a valid jury waiver. To begin with, petitioner's medical records indicate that he tended to exaggerate his retardation in a manipulative fashion. This court is persuaded by the medical record that petitioner was only mildly retarded at the time of trial, and that such retardation did not substantially impair his ability to understand either his criminal proceedings or his jury waiver. At least arguably, counsel's failure to make a formal recommendation cuts against petitioner in that there was no effort to unduly control petitioner's will. Ciummei, supra, at 1192, n. 13. As for the failure to render a complete description of the jury's nature and function, the court agrees that the better practice one now imposed on Massachusetts judges would have been to give such a description by means of a contemporaneous colloquy. But that failure does not affect the validity of petitioner's waiver, given the other indicia of voluntariness and knowledge. Finally, while the master detected some confusion in petitioner's knowledge of the jury process, this again is not fatal in the light of the overall record, including his previous jury waivers and his past experience with the criminal justice system. After an independent evaluation of the record, the court concludes that petitioner's jury waiver satisfied constitutional criteria. *942 III. Petitioner's third argument concerns the refusal of the Supreme Judicial Court to apply the colloquy requirement announced in Ciummei to Ciummei himself. He argues that the "completely prospective" nature of that requirement violates equal protection in that it affects a fundamental interest (the right to a jury trial) while serving no compelling state interest sufficient to withstand "strict scrutiny." As support for this argument, petitioner cites the Supreme Court's treatment of then pending cases raising the same issue as that resolved in Miranda v. Arizona, 384 U.S. 436, 86 S.Ct. 1602, 16 L.Ed.2d 694 (1966). The holding in Miranda was prospective, except as to Miranda himself and three other petitioners whose cases had been consolidated for argument. All other pending cases raising the same issue were dismissed without relief. See Desist v. United States, 394 U.S. 244, 255, 89 S.Ct. 1030, 1037, 22 L.Ed.2d 248 (1969) (Douglas, J., dissenting). Petitioner argues that he is entitled to retroactive relief afforded Miranda and his three co-petitioners. His point is that such an approach serves as incentive for litigants to raise law reform issues by holding out the hope of individual relief, even in cases where the reform is generally applied prospectively. Were petitioner's argument cast solely in terms of traditional retroactivity doctrine, this court would not hesitate to reach its merits. The Supreme Judicial Court in Ciummei explicitly considered some of the factors relevant to that doctrine in deciding to make the colloquy requirement wholly prospective. Ciummei, supra, at 1190 & n. 10. But there is nothing on the record to indicate that the Supreme Judicial Court had before it petitioner's contention that the court's refusal to accord him individual relief would amount to a violation of equal protection. A habeas corpus petitioner must exhaust state remedies prior to seeking federal relief. There has been no such exhaustion with respect to petitioner's equal protection argument. Petition dismissed. An order will issue. NOTES [1] For cases rejecting constitutionalization of the colloquy, see Estrada v. United States, 457 F.2d 255 (7th Cir.), cert. denied, 409 U.S. 858, 93 S.Ct. 143, 34 L.Ed.2d 104 (1972); United States v. Mitchell, 427 F.2d 1280 (3d Cir. 1970); United States v. Straite, 425 F.2d 594 (D.C.Cir. 1970); United States v. Hunt, 413 F.2d 983 (4th Cir. 1969); State v. Jelks, 105 Ariz. 175, 461 P.2d 473 (1969); Quartz v. State, 258 So.2d 283 (Fla.Dist.Ct.App.1972); State v. Olivera, 53 Haw. 551, 497 P.2d 1360 (1972); Kindle v. State, 161 Ind.App. 14, 313 N.E.2d 721 (1974). In Walker v. State, 578 P.2d 1388, 1390 (Alaska 1978), it is unclear whether the court based its requirement of a colloquy on state or federal constitutional law or both. In Short v. Commonwealth, 519 S.W.2d 828, 833 (Ky. 1975), the court applied the guilty plea procedure to a jury waiver case on state constitutional grounds. [2] Independent of any constitutional requirement, several jurisdictions have deemed it advisable to require a contemporaneous colloquy by statute, rule, or in the exercise of a general supervisory power. See, e. g., United States v. Scott, 583 F.2d 362, 363-64 (7th Cir. 1978) (supervisory power); see also Ciummei supra, at 1189 (citing cases). Indeed, the Supreme Judicial Court took the opportunity in Ciummei to require, "in aid of sound judicial administration," the holding of an in-court colloquy to test the knowledge and willingness of the defendant offering to waive the jury right. Ciummei, supra, at 1189. That requirement was made wholly prospective. Id. at 1190. [3] Under Mass.Gen.Laws ch. 263, § 6, and Mass. R.Crim.P. 19(a), a defendant may waive his jury right by submitting a written waiver subject to court approval. The requirement of court approval was added in 1979. In the federal system, jury waiver requires the concerted consent of the government, the defendant, and the court. See Fed.R.Crim.P. 23(a). [4] Petitioner has not argued that the trial judge's determination of competence to stand trial was erroneous. See Ciummei, supra, at 1192-93 n. 15. [5] Section 2254(d) lists eight factors which, if present, remove the presumption of conclusiveness to be accorded state court findings. Those factors relate to the adequacy of the state court fact finding procedures, the jurisdiction of the state court, the precise issues decided, the presence of counsel for the defendant, and any due process violations that may have occurred. [6] Neither party has attacked the validity or conclusiveness of those factual findings. [7] Plaintiff ultimately was convicted of unarmed robbery. | Low | [
0.48314606741573,
32.25,
34.5
] |
Research and timeline by Djordje Djukic; map and additional reporting by Evan Centanni Though Iraq's current war has been proceeding at a less-intense pace than Syria's, there have been several important changes to territorial control in the past half a year, all involving the so-called "Islamic State" (ISIS) losing ground to the Iraqi government and Kurdistan's Peshmerga force.See all this and more on the newest update to PolGeoNow's concise, professional-quality Iraq control map, which includes a timeline of changes since our previous Iraq map report in July. This map and report are professional subscriber content, accessible to paid members or by separate purchase. Download PDF (US$19.99) | Low | [
0.46201232032854206,
28.125,
32.75
] |
Total Synthesis of (-)-Exiguolide. A concise total synthesis of (-)-exiguolide has been completed in an overall 2.8% yield over 20 steps in the longest linear path. The key strategies involve (1) Prins cyclization/homobromination of dienyl alcohol with the B ring-substituted aldehyde, prepared by Prins cyclization/bromination, to construct the A ring with excellent cis-Z stereochemical control and (2) an unusual side chain installation/macrocyclization strategy featuring Sonogashira cross-coupling followed by a ring-closing metathesis reaction to deliver the target. | High | [
0.677871148459383,
30.25,
14.375
] |
Effects of immersion in cholesterol-lipid solution on the tensile and fatigue properties of elastomeric polymers for blood pump applications. Tensile and fatigue properties were studied on three kinds of segmented polyether polyurethanes developed for blood pump applications. To study the effects of plasma constituents, specimens were immersed in a cholesterol-lipid solution whose composition is similar to that of plasma. The results obtained were compared with those observed in a saline solution which had been reported previously. Immersion in the cholesterol-lipid solution has influence on the mechanical properties of Toyobo TM5 polyurethane and Avcothane 51: Its effect is enhanced by cyclic deformation. Cholesterol-lipids change the mechanical characteristics of Avcothane 51 more remarkably than those of Toyobo TM5 polyurethane possibly because the former contains about 10% silicone. On the other hand, there is little effect of the constituents of the cholesterol-lipid solution on the static and dynamic mechanical properties of Biomer, although moisture and cyclic deformation have some influence on the mechanical characteristics. | Mid | [
0.64390243902439,
33,
18.25
] |
Neonatal abstinence syndrome and cerebral infarction following maternal codeine use during pregnancy. Neonatal withdrawal from maternal drugs and medications is common in some NICUs. Codeine-containing cough preparations given to pregnant mothers have been identified as a cause of neonatal abstinence syndrome. However, many women do not consider prescription cough syrups when asked about drug use. Maternal medication or illicit drug use has been identified as a cause of perinatal arterial stroke. Since codeine is an opiate with similar pharmacodynamic effects to morphine, it is reasonable to investigate if maternal codeine use has effects on the fetus that are similar to other opiates. The authors present 2 cases of newborn infants with perinatal arterial stroke that may have been associated with in utero exposure to codeine. Physicians should ask about maternal medication use, including codeine-containing cough preparations, when evaluating newborn infants with evidence of cerebral infarction. | High | [
0.668280871670702,
34.5,
17.125
] |
This year's fourth issue of Shonengahosha's Young King magazine (issue cover pictured at right) announced on Monday that Boichi will publish the final chapter of the Sun-Ken Rock manga in two more issues. The manga will take a break in the next issue, and then end in this year's sixth issue of Young King on February 22. The series tells the story of Japanese teenager Ken Kitano, who follows the girl he likes to Seoul, Korea. After he arrives, his impressive fighting skills and strong spirit draw the attention of a gang. Boichi had launched the final arc in the series in August 2013. The 24th compiled book volume teased in October that the manga will end in the 25th volume. The manga has been running in Shonengahosha's Young King magazine since 2006. JManga released part of the series in English online starting in July 2012 before the site ended its service in May 2013. Crunchyroll now releases the manga online in English. Boichi launched a spinoff manga of Yū Sasuga and Kenichi Tachibana's Terra Formars manga titled Terra Formars Asimov (Terra Formars Gaiden Asimov) in the 23rd issue of Shueisha's Grand Jump magazine in November. | Mid | [
0.630669546436285,
36.5,
21.375
] |
LEGO Indiana Jones demo on Live Some combination of Activision, LucasArts, Traveller's Tales and Microsoft have plonked a demo of LEGO Indiana Jones on Xbox Live Marketplace. Which is a bit odd, because Activision told us just last month that there wouldn't be! Released earlier this month to general acclaim, LEGO Jones explores the original three films in the same way the LEGO Star Wars games caricatured their inspiration. Also like LSW, LEGO Jones adapts key elements of the film for its LEGO-tinged gameplay mechanics: Indy does some basic archaeology to solve puzzles, punches stuff and swings and climbs on ropes. We ended up rather liking it (Dan Whitehead called it "a commendable improvement on an already solid framework") but also felt it was a shame that many of the extra bits that made the LSW games so deep and varied were absent. There's no word on how big the demo download is but it should be available everywhere. | Mid | [
0.5507246376811591,
33.25,
27.125
] |
{"type":"FeatureCollection","properties":{"kind":"state","state":"TX"},"features":[ {"type":"Feature","properties":{"kind":"county","name":"Wood","state":"TX"},"geometry":{"type":"MultiPolygon","coordinates":[[[[-95.1530,33.0131],[-95.1530,32.9035],[-95.1530,32.5694],[-95.1804,32.5420],[-95.3447,32.6078],[-95.4761,32.6078],[-95.5747,32.6735],[-95.5966,32.6844],[-95.6350,32.7228],[-95.6678,32.9583],[-95.3063,32.9638],[-95.1749,32.9638]]]]}} ]} | Low | [
0.502958579881656,
31.875,
31.5
] |
Q: Circle with CSS HTML: <ul class="list_header"> <li><a href="#">06 december 2013</a></li> <li><a href="#">item2</a></li> <li><a href="#">item3</a></li> <li><a href="#">item4</a></li> <li><a href="#">item5</a></li> <li><a href="#">item6</a></li> </ul> CSS: .list_header{ font: italic 70% Georgia; margin: 15px 0 0 15px; } .list_header li{ display: inline-block; } .list_header a{ color: #bbb; text-decoration: none; } .list_header li:first-child :after{ content: '/t'; height: 3px; border-radius: 50%; background: #bbb; padding: 0px 0px 0px 5px; margin: 0px 0px 0px 5px; } http://codepen.io/hristofor/pen/aILlC How to made this circle smaller? I tried to make the circle smaller with width and height but it didn't work. What to do? A: First remove the tab character from the content, then specify display: inline-block, remove the padding, and set width and height to the same desired value. Live example: http://codepen.io/anon/pen/Iyeth The problem in your code is that the pseudo element is inline, so the width and the height depend from its content and other properties like font-size. | Low | [
0.5269320843091331,
28.125,
25.25
] |
Sunday, 22 September 2013 Virtual gardens illuminate real-world attitudes about nature Researchers have long struggled to design surveys that collect detailed and informative data without introducing bias through the use of loaded, confusing, or restrictive wording. A team of French researchers has come up with a novel solution to this problem: Toss out the surveys altogether, and replace them with a virtual computer program that allows respondents to express their thoughts and preferences in actions rather than in words. This is the idea behind Virtual Garden, a program that allows users to select from 95 different features in order to create their ideal garden. The features fall within seven different categories: animals, flowers, lawn and cover, sport and playing, trees and bushes, water, and other. The program keeps track of each item that is added and adjusted, and calculates biodiversity as biotic features are introduced into the virtual habitat. Users are able to view the garden from multiple angles and even take a virtual stroll through the area in order to evaluate their progress and determine whether further manipulations need to be made to the environment. Far from being merely a nerdy new version of The Sims, the program was designed to assess which features people most want to experience when they visit public gardens. Further, an analysis of the virtual habitats could help clarify the role that biodiversity has in driving humans' overwhelmingly positive responses to green spaces--particularly those located in otherwise urban areas. Finally, by collecting basic social, economic, and demographic information about each program user, the program's developers can also assess whether habitat preferences are influenced by age, education, income, and general interest in the natural environment. The research team responsible for Virtual Garden trialled the program among 732 Parisian hospital patients. Each individual was given a 30-minute time limit for designing the garden, though the average length of time required was only 19.2 minutes. Gardens typically contained approximately 24 different features--9 "objects" (such as ponds), 5 animals, 8 flowers, and 5 woody species (trees or bushes). Overall, users included fewer biotic features than were expected by chance. Animals were particularly underrepresented, with nearly a third of gardens containing no animals at all, and almost another third containing fewer than 5 animal species. Larger animals--especially mammals and herptiles--were not very popular; the least preferred species overall were foxes and chimpanzees. Ladybugs, peacocks, and great tits, on the other hand, were the most preferred. The most popular species were generally those that are common in Parisian gardens, suggesting that patients tended to populate their virtual gardens with species that are most familiar to them. Several demographic and socioeconomic factors influenced garden design. For example, men included fewer animals and flowers than women; younger patients included more non-native species; and people who showed a greater interest in conservation and nature activities tended to create gardens with higher biodiveristy. Interestingly, plant richness was higher in gardens created by people who grew up in more rural areas, again suggesting that familiarity with species is an important driver of habitat preferences. The Virtual Garden trial produced two main results. First, it suggests that computer programs may be a useful way to collect data from people without accidentally introducing bias into a study. Such programs are likely to be particularly useful in situations where researchers need to address or describe situations that are highly visual in nature--such as habitat structure, the aesthetics of which can greatly influence respondents' attitudes and opinions. Second, the patterns reported here get us one step closer to understanding city-dwellers' complex and often contradictory responses to green spaces. There is particularly strong evidence of an "extinction-of-experience" process, whereby people judge biodiversity and aesthetics according to what they have previously experienced, rather than what may be natural, healthy, and/or desirable in a given environment. The creators of the Virtual Garden hope that conservationists and managers can use their program to collect and compare data from across a wide geographic range, and, therefore, to improve our "understanding of the role culture and living context...play in people's relations with biodiversity." This could not only help save threatened species, but also improve the well-being of people by increasing and improving human-nature interactions even in the most urban of environments. 1 comment: Your blog article is very nice thank you for share this information students are getting certificates for their academic life now colleges are taken taken by the additional courses after completing that courses. Who is the "Anthrophysist"? I am a biologist who studies the ways in which anthropogenic disturbance impacts animals (especially birds). I hope that the results of my work, and the work of other researchers like me, can help humans learn how to coexist more peacefully with wildlife. I am also interested in the role that nature has played in shaping human cultures around the world and over the centuries. Although this blog will predominantly focus on scientific research, I hope to occasionally profile some anthropological work as well, in order to better highlight the interconnectedness of humans ("anthro") and nature ("physis"). | Mid | [
0.600896860986547,
33.5,
22.25
] |
Did you see the story in The News about the mother in a row with a Fareham school over her daughter using an eyebrow pencil? Hannah Knight, 11, was ordered to remove the make-up and her mother Karen has since kept her out of school until matters are resolved. When children are made to wear the same outfit day after day, told to comply, sometimes we all need to take a step back and remember the perils of our own childhood and how we responded to them This story came up in my news feed on social media and I was surprised by the comments, which were aghast at the mother’s behaviour. It is very easy to sit on the sidelines and make judgment calls about who should be doing what and their motives. But, actually, there is often a lot more going on of which us, the commentators, are unaware. I worked in a senior school for a few years and have witnessed young ladies wearing far more make-up in a single day than I have cumulatively over my entire life. But here’s the thing. I’ve never been plagued by acne, or spots. Yes, I’ve had one or two, but haven’t had the debilitating situation where someone is so upset by the way they look, they have to hide away. I have had it with my hair though. It’s unruly and it takes a brave hairdresser to get involved with the curls, especially when they’re wiry, tight corkscrews at the front and relaxed at the back. I’ve had more than one dodgy result in the past and have, I admit, taken time off work to tame my hair’s worst points with scarves and slides. And while people can tell you that no-one notices, whatever your Achilles heel is, you know how it feels when it’s not at its best. As adults we invest our time in a certain style of dress, or a certain hairdo, to make ourselves feel better. So I’m asking that we put the lid back on the judgments and instead think about how it feels to do whatever it is to make yourself feel confident. When children are made to wear the same outfit day after day, told to comply, sometimes we all need to take a step back and remember the perils of our own childhood and how we responded to them. I’m not judging anybody. I’m just saying that people in glass houses should be very careful with stones. LOOKING FOR A CRACKING BEACH READ? TRY YOUNG ADULT FICTION Every now and then I remember how much I love the library and my reading is invigorated. My secret is focusing on young adult fiction as authors for young adults tackle really interesting subject matter and themes in excellent shorter formats than some adult authors (and also I’m published in that genre). If you fancy a few great beach reads, do try the teen-plus range. It’s not all Dystopian futures based around killing one another. There are plenty that speak to all sorts of issues and stories and characters. Teen fiction is concise, accessible and digestible. With summer finally here, it’s worth remembering that your library can take you to places that your finances can’t. AFTER BEING YELLED AT, MAYBE IT’S TIME TO LOSE THE LANDLINE I was yelled at by a cold caller the other day. Can you believe it? Apparently it was me who was rude by asking ‘what?’ and ‘why?’ Not him for calling me up in the middle of the day and demanding to speak to my husband. Not him for insinuating that I was being ruder and ruder. Yes, I understand that it’s his job to ring people up. But to then yell at me and tell me off? I was less than pleased, then I was amused and now I’m simply peeved that my phone number is on some list somewhere and however many times I try to get myself removed, I simply can’t do it. Perhaps, in the age of the mobile, it’s time to get rid of the landline. | Low | [
0.5328947368421051,
30.375,
26.625
] |
Card Holder Activity Tray is a great way to enjoy playing cards again. The R.O.S. Card Holder Board was designed to make it easier for someone to hold playing cards just in case their hands and fingers aren't what they used to be. The larger board can hold up to 21 over-sized playing cards and the smaller board can hold up to 14 standard sized playing cards. This can be used as a stand alone card holder or as a top to the R.O.S. Activity Console. High density material is dishwater safe and can be cleaned with soap and water, Formula 409, sanitary wipes, Fantastik, etc. This supplier accepts return of new, unopened items in their original packaging, including manuals, warranties, etc. in re-saleable condition within 30 days of the original ship date. A restock fee of 25% applies to all returns, unless the item is defective. No returns accepted on items more than 30 days after original ship date. Stay connected with Parentgiving. New subscribers will save 10% on your first order!* *New subscribers only. This coupon cannot be combined with other offers, used on Dry Direct orders or used on Golden Technologies or Pride Mobility products. Maximum savings of $75. Some restrictions apply. | Mid | [
0.5565862708719851,
37.5,
29.875
] |
That's the idea saladdodger, but my triangle is so small and down sloping that I need one that sits on top of my top tube. I've found that I can happily fit a 5L dry bag in an anything cage seated in the triangle. Long Top tube bag would be good for tarp poles etc | Mid | [
0.6370558375634511,
31.375,
17.875
] |
# #=BEGIN SONGBIRD GPL # # This file is part of the Songbird web player. # # Copyright(c) 2005-2010 POTI, Inc. # http://www.songbirdnest.com # # This file may be licensed under the terms of of the # GNU General Public License Version 2 (the ``GPL''). # # Software distributed under the License is distributed # on an ``AS IS'' basis, WITHOUT WARRANTY OF ANY KIND, either # express or implied. See the GPL for the specific language # governing rights and limitations. # # You should have received a copy of the GPL along with this # program. If not, go to http://www.gnu.org/licenses/gpl.html # or write to the Free Software Foundation, Inc., # 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. # #=END SONGBIRD GPL # DEPTH = ../../../.. topsrcdir = @top_srcdir@ srcdir = @srcdir@ VPATH = @srcdir@ include $(DEPTH)/build/autodefs.mk SONGBIRD_TEST_COMPONENT = library-base SONGBIRD_TESTS = $(srcdir)/head_library-base.js \ $(srcdir)/test_constraints.js \ $(srcdir)/test_dynamicplaylist.js \ $(srcdir)/test_librarymanager.js \ $(srcdir)/test_timeformatter.js \ $(NULL) # TODO: XXX Need to fix these tests and then add them back in # see bugs 23940-23945 # $(srcdir)/test_mediaitemwatcher.js \ # $(srcdir)/test_removalmonitor.js \ # $(srcdir)/test_media_item_controller_cleanup.js \ include $(topsrcdir)/build/rules.mk | Mid | [
0.6007604562737641,
39.5,
26.25
] |
divisor of 121 and 99. 11 Calculate the highest common divisor of 1696 and 224. 32 What is the highest common factor of 17 and 1513? 17 What is the greatest common factor of 2552 and 88? 88 Calculate the greatest common factor of 5 and 6655. 5 What is the highest common factor of 27 and 531? 9 What is the greatest common factor of 583 and 209? 11 What is the highest common factor of 35 and 55? 5 What is the highest common divisor of 55 and 16445? 55 What is the highest common factor of 31 and 1519? 31 What is the highest common factor of 30 and 18? 6 Calculate the greatest common divisor of 93 and 22413. 93 Calculate the highest common factor of 2451 and 684. 57 What is the highest common factor of 34 and 119? 17 What is the highest common divisor of 176 and 20? 4 What is the greatest common factor of 41 and 656? 41 What is the greatest common factor of 933 and 3? 3 Calculate the highest common factor of 5 and 20. 5 Calculate the greatest common factor of 1582 and 14. 14 Calculate the greatest common factor of 945 and 35. 35 What is the greatest common factor of 27 and 36? 9 Calculate the highest common divisor of 192 and 6288. 48 What is the greatest common divisor of 3440 and 5? 5 Calculate the highest common divisor of 2583 and 41. 41 What is the greatest common divisor of 63 and 189? 63 What is the greatest common divisor of 54 and 81? 27 What is the greatest common divisor of 2470 and 260? 130 Calculate the highest common factor of 1457 and 496. 31 Calculate the highest common divisor of 7021 and 2023. 119 What is the highest common divisor of 2500 and 1200? 100 Calculate the highest common factor of 3243 and 1311. 69 Calculate the highest common factor of 12 and 66. 6 What is the greatest common divisor of 14 and 35? 7 Calculate the highest common divisor of 8 and 536. 8 Calculate the greatest common factor of 107476 and 388. 388 Calculate the greatest common divisor of 108 and 4. 4 Calculate the greatest common factor of 8877 and 11. 11 What is the greatest common divisor of 16 and 1752? 8 What is the highest common factor of 507 and 3? 3 What is the highest common factor of 7390 and 15? 5 Calculate the greatest common factor of 13 and 1027. 13 Calculate the greatest common factor of 6994 and 104. 26 Calculate the highest common factor of 828 and 468. 36 Calculate the highest common divisor of 376 and 392. 8 What is the greatest common factor of 148 and 7659? 37 Calculate the highest common divisor of 264 and 88. 88 What is the greatest common divisor of 10 and 675? 5 What is the greatest common divisor of 48 and 5344? 16 Calculate the greatest common divisor of 54 and 12. 6 Calculate the highest common factor of 437 and 1235. 19 Calculate the greatest common divisor of 3842 and 17. 17 What is the greatest common divisor of 1400 and 50? 50 Calculate the highest common factor of 18 and 1698. 6 Calculate the greatest common divisor of 533 and 13. 13 Calculate the greatest common divisor of 378 and 36. 18 What is the highest common divisor of 844 and 8? 4 Calculate the greatest common divisor of 19208 and 24. 8 Calculate the highest common factor of 240 and 80. 80 Calculate the greatest common factor of 28 and 4298. 14 What is the highest common divisor of 72 and 2724? 12 Calculate the greatest common factor of 592 and 9. 1 What is the highest common divisor of 1000 and 40? 40 Calculate the greatest common factor of 165 and 3585. 15 Calculate the greatest common factor of 26 and 7046. 26 Calculate the highest common factor of 6 and 1107. 3 What is the highest common divisor of 204 and 96? 12 Calculate the highest common factor of 836 and 44. 44 Calculate the greatest common divisor of 61 and 46421. 61 What is the greatest common divisor of 15 and 411? 3 What is the highest common divisor of 576 and 216? 72 Calculate the greatest common factor of 4141 and 123. 41 Calculate the greatest common divisor of 72 and 88. 8 What is the highest common factor of 48 and 216? 24 What is the highest common divisor of 153 and 34? 17 What is the highest common divisor of 684 and 90? 18 What is the greatest common factor of 48 and 32? 16 Calculate the greatest common factor of 51 and 374. 17 Calculate the greatest common divisor of 30435 and 15. 15 What is the greatest common divisor of 19 and 2831? 19 Calculate the highest common factor of 33 and 165. 33 Calculate the greatest common divisor of 230 and 85. 5 Calculate the greatest common factor of 182 and 210. 14 What is the highest common divisor of 12 and 372? 12 Calculate the greatest common factor of 65 and 13. 13 What is the highest common divisor of 6 and 12126? 6 What is the greatest common divisor of 340 and 289? 17 What is the greatest common factor of 285 and 1501? 19 Calculate the highest common factor of 30 and 15. 15 Calculate the highest common divisor of 432 and 16. 16 What is the highest common factor of 288 and 32? 32 What is the greatest common factor of 11067 and 17? 17 Calculate the greatest common divisor of 3102 and 132. 66 What is the greatest common factor of 94 and 2? 2 Calculate the highest common factor of 118 and 3186. 118 Calculate the highest common divisor of 117 and 312. 39 What is the highest common factor of 7 and 63? 7 What is the greatest common factor of 6248 and 11? 11 What is the highest common factor of 1133 and 11? 11 Calculate the highest common factor of 150 and 330. 30 Calculate the highest common divisor of 28 and 98. 14 Calculate the highest common divisor of 1422 and 3634. 158 Calculate the highest common divisor of 84 and 51. 3 What is the greatest common divisor of 36 and 156? 12 What is the highest common factor of 297 and 54? 27 What is the highest common divisor of 10720 and 480? 160 Calculate the highest common divisor of 14 and 1106. 14 Calculate the highest common divisor of 47453 and 7. 7 What is the greatest common divisor of 1056 and 1155? 33 Calculate the highest common divisor of 4963 and 7. 7 What is the highest common divisor of 324 and 567? 81 Calculate the greatest common divisor of 660 and 180. 60 Calculate the highest common factor of 21 and 19005. 21 Calculate the greatest common factor of 54 and 19494. 54 Calculate the greatest common factor of 98 and 5586. 98 What is the highest common factor of 75 and 24? 3 Calculate the highest common factor of 480 and 2. 2 Calculate the highest common factor of 25 and 7075. 25 Calculate the highest common factor of 87 and 928. 29 Calculate the greatest common factor of 64 and 16064. 64 What is the highest common divisor of 18 and 120? 6 Calculate the greatest common factor of 48 and 84. 12 What is the highest common factor of 361 and 2299? 19 Calculate the greatest common divisor of 3 and 2319. 3 What is the greatest common factor of 1592 and 104? 8 Calculate the highest common divisor of 276 and 60. 12 What is the highest common factor of 1460 and 580? 20 Calculate the greatest common divisor of 569 and 1. 1 Calculate the highest common divisor of 91 and 65. 13 Calculate the highest common factor of 13 and 27677. 13 What is the greatest common factor of 33 and 187? 11 What is the greatest common divisor of 8 and 696? 8 What is the highest common factor of 188 and 893? 47 Calculate the highest common divisor of 7 and 623. 7 Calculate the highest common factor of 1260 and 54. 18 What is the highest common factor of 206 and 6386? 206 What is the highest common factor of 22 and 319? 11 Calculate the highest common factor of 232 and 3944. 232 Calculate the highest common divisor of 9 and 59229. 9 Calculate the greatest common divisor of 126 and 414. 18 What is the greatest common factor of 432 and 270? 54 Calculate the highest common factor of 162 and 1854. 18 What is the highest common divisor of 35 and 10955? 35 Calculate the highest common factor of 1216 and 76. 76 Calculate the highest common factor of 5684 and 116. 116 What is the greatest common divisor of 7 and 2? 1 What is the greatest common divisor of 3930 and 30? 30 What is the highest common divisor of 1534 and 104? 26 What is the highest common factor of 9493 and 55? 11 Calculate the greatest common divisor of 369 and 36. 9 What is the greatest common factor of 2048 and 544? 32 What is the greatest common divisor of 396 and 636? 12 Calc | Low | [
0.5373737373737371,
33.25,
28.625
] |
India Transparency International corruption index blow Published duration 1 December 2011 image caption Anti-corruption protests have gathered pace in India India has suffered a new blow to its reputation on corruption, dropping eight places on a key annual list by a leading anti-corruption watchdog. The Transparency International Corruption Perceptions list ranks a total of 183 nations from a best of 10 to a worst of 0. India fell from a 3.3 ranking in 2010 to 3.1, dropping from 87th to 95th. The Indian government has been hit by a number of corruption scandals with a major new bill before parliament. Parliament's winter session began on 22 November with discussion of the landmark anti-corruption law a priority. However, there has been uproar in the lower house over a number of issues and virtually no legislative work has been done. Ombudsman India ranks lower than China (75th) on the new list but higher than Pakistan (134th). Nepal at 154th is the lowest ranking South Asian nation. New Zealand is top and Somalia bottom of the list. Almost two-thirds of the nations scored less than five, with Transparency International saying: "Corruption continues to plague too many countries around the world." The Indian government is under pressure from anti-corruption campaigners to pass the new law to allow the setting up of a citizens' ombudsman, also known as the Jan Lokpal. Leading anti-corruption campaigner Anna Hazare, who says he has the support of millions, has warned he will resume a hunger strike and revive street protests that rocked the government this year if the new law does not come into force. Mr Hazare went on a 12-day hunger strike in August. | High | [
0.6621315192743761,
36.5,
18.625
] |
Determination of Multimodal Isotopic Distributions: The Case of a (15)N Labeled Protein Produced into Hairy Roots. Isotopic labeling is widely used in various fields like proteomics, metabolomics, fluxomics, as well as in NMR structural studies, but it requires an efficient determination of the isotopic enrichment. Mass spectrometry is the method of choice for such analysis. However, when complex expression systems like hairy roots are used for production, multiple populations of labeled proteins may be obtained. If the isotopic incorporation determination is actually well-known for unimodal distributions, the multimodal distributions have scarcely been investigated. Actually, only a few approaches allow the determination of the different labeled population proportions from multimodal distributions. Furthermore, they cannot be used when the number of the populations and their respective isotope ratios are unknown. The present study implements a new strategy to measure the (15)N labeled populations inside a multimodal distribution knowing only the peptide sequence and peak intensities from mass spectrometry analyses. Noteworthy, it could be applied to other elements, like carbon and hydrogen, and extended to a larger range of biomolecules. | High | [
0.672941176470588,
35.75,
17.375
] |
West Lake Sammamish Parkway is set to reopen by mid-April, but a city utility project may impact traffic along 164th Avenue beginning March 26. The city will replace aging water pipes along 164th Avenue between Northeast 8th and Southeast 14th streets from March 26 through June or July. To minimize traffic disruptions on 164th Avenue the first two weeks, the contractor will be doing work in the east shoulder of 164th Avenue, between SE 9th and SE 14 Streets. This will allow continuous two-way traffic along 164th Avenue with only ccasional interruptions. Work will take place Monday through Friday, 7:30 a.m. to 4 p.m. Traffic impacts are likely the third week of construction when 164h Avenue will still be open to traffic, but will have flaggers dropping traffic to one lane at the location of the work crew. Drivers will be advised to use other roads. See entire City map for possible alternate routes. Electronic message boards have been in place for a couple of weeks, advising drivers of the utility project on 164th. During the week of March 26, standard construction detour signs will be placed and remain for the duration of the construction work. While the detour change may cause some inconvenience until West Lake Sammamish Parkway reopens, it's important that the city's utility work begin this month to stay on schedule and be completed ahead of a city overlay project, which will begin later this summer on 164th Avenue. Residents will be informed about pavement resurfacing construction start and end dates later this summer. West Lake Sammamish Parkway has been closed since Jan. 22 due to a mudslide that undercut the roadway. When the steep slope gave way on the east side of the Parkway, dirt and debris under the roadway slid down the hill, leaving an unsupported section of concrete panels. With the redesign complete and the necessary property easements in place, construction began the first week in March. | Mid | [
0.543396226415094,
36,
30.25
] |
Topic of the program: What kind of examination do the laboratory- instrumental examinations involve, and is there any standard set by the ministry about what is considered to be the high technological examination? Guest of the program: Rusudan Rukhadze- Head of the Department of Health at Ministry of Health of Georgia | Mid | [
0.6366120218579231,
29.125,
16.625
] |
Q: What does Tate mean when he wrote "Higher dimensional class field theory" in the new preface to the Artin-Tate book and another question? This is probably well-known to the experts or many number theory students, but since I am just starting to learn class field theory (with some basic knowledge of algebraic numbers, e.g. the 3 basic theorems on Dedekind domains, class groups, units; and a little on local/global fields), I would really love to learn what Tate means when he wrote "Higher dimensional class field theory" in the new preface to the Artin-Tate book, as a new direction to the classical class field theory that deals with abelian extensions. So, what does he mean by that? I'd appreciate a summary, and also references/books/papers if necessary. I can sort of understand that the "usual" class field theory is probably about "1-dimensional objects", made precise by Dedekind's theorem that a prime in a Dedekind domain can be factorized into primes that are determined by a single polynomial (the primes being in the integral closure of the Dedekind domain in a finite separable extension), under certain conditions. Although this is well-known, for definiteness, it's on p. 16 of Professor James Milne's notes on Class Field Theory. I am not sure how to generalize this to higher dimensions; neither am I sure that the generalization of this theorem answers my own question of what Tate means (and what are "higher dimensional fields" anyway?) I can think of yet another potential meaning of this: from a popular account of what goes into the proof of Fermat's Theorem (the book "Fearless Symmetry"), the quadratic reciprocity law is interpreted as an result on the 1-dimensional Galois representation (I only know the definition of this), so maybe using higher-dimensional Galois representations, maybe one can generalize the class field theory to "higher dimensions". Alas, all these are my speculations and fantasies, probably due to my ignorance on the subject (starting to learn CFT); but I really like to put everything in as nice a perspective (or perspectives) as possible while I am learning CFT. So this leads naturally to my other question below: How do I find an "optimal" and yet pedagogical (to me) way to learn CFT, incorporating the perspectives and preparations for the most important modern developments in number theory (and potentially algebraic/arithmetic geometry, representation theory, Langlands, etc.)? Are there nice references/books/papers along the way of learning CFT that I should be also looking at? Thank you very much for your kind suggestions and help. (EDIT) Thanks to Matt E's explanation of the 2 terms. The 2nd and 3rd to last paragraph of his "Preface to the New Edition" (following link http://books.google.com/books?id=8odbx9-9HBMC&lpg=PP1&dq=class%20field%20theory&pg=PR6#v=onepage&q&f=false) are the relevant passages, where he also wrote down a few other examples of new directions. I only ventured to try to understand the first one, and has a very very rough superficial idea what the 2nd one is generally about - this means I will come back for more of these after I learn more! Since these 2 background passages are a bit long (esp. 3rd to last one, which might give a better context for what Tate means), so please excuse me for providing a link. Please let me know if you can't see the paragraphs from the preview. A: By "higher dimensonal class field theory", Tate means the class field theory of higher dimensional local fields (see also this brief discussion), developed in the work of various people, including Kato and Parshin. As for your second question, about learning CFT from a modern perspective: with my own students, I encourage them to learn from Cassels and Frolich (including the exercises), from Cox's book Primes of the form $x^2 + n y^2$, and from Washington's article on Galois cohomology in Cornell, Silverman, and Stevens. The first reference (especially the main articles of Serre and Tate) gives a development of the main results of CFT which I think is hard to beat. Cox's book gives an important classical perspective. Washington's article gives insight into how class field theory can be reformulated as a collection of theorems (mainly due to Tate) on (local and global) Galois cohomology. Tate's article Number theoretic background in the second volume of Corvalis is good when you have reached a certain level of sophistication, and are ready to move on from just focussing on algebraic number theory and CFT to a broader perspective. My experience is that it is a little austere for a beginner, though. One thing that you will be missing if you follow the above references is an $L$-function-based perspective on class field theory. I gather that this is discussed in the new edition of Artin--Tate. If so, it is worth learning, since although it is the more old-fashioned point of view on CFT, non-abelian class field theory (i.e. the Langlands program) is founded on the notion of $L$-functions. (I believe that Lang's book also discusses the $L$-function approach to CFT, but I've never read it myself.) | Mid | [
0.6284987277353691,
30.875,
18.25
] |
{ "images" : [ { "idiom" : "universal", "filename" : "wuhantong.jpeg", "scale" : "1x" }, { "idiom" : "universal", "scale" : "2x" }, { "idiom" : "universal", "scale" : "3x" } ], "info" : { "version" : 1, "author" : "xcode" } } | Low | [
0.44743276283618505,
22.875,
28.25
] |
Q: Spring Boot "Ignored because not a concrete top-level class" I have this (https://github.com/Danix43/HerculesAPI) REST application made by Java with Spring Boot. All good when for some reason the @SpringBootApplication annotation doesn't find anymore the interfaces TermometruService annothed with @Service. I have tried everything from using @ComponentScan with the service package but then any Rest operation returns a RESOURCE_NOT_FOUND to changing the jdk version and a flush of maven repositories. Project structure if useful The log after running the application 2020-01-27 19:33:39.635 DEBUG 6248 --- [main] o.s.c.a.ClassPathBeanDefinitionScanner : Ignored because not a concrete top-level class: file [D:\Programe\Programare\Java\HerculesAPI\target\classes\com\herculesapi\services\TermometruService.class] A: The problem was solved by annotating the TermometruServiceImple class with @Service. Also regarding the Github repository, it doesn't always have all the changes right away pushed but yes the TermometruService interface had the annotation. | Mid | [
0.65079365079365,
30.75,
16.5
] |
Air Free Onix 3000 Air Sterilizer With renovation dust in the air at IMPACT’s new HQ in Calgary, we plugged in the Air Free Air Sterilizer to clear the proverbial air. Now, we have no science to back up our experience, but the collective opinion is that we are breathing easier here. The company cites independent lab studies that Air Free eliminates 70 to 99 per cent of airborne microorganisms such as fungus, bacteria and dust mite allergens. This little Darth Vader lookalike is a silent killer, incinerating contaminants in its ceramic heating core. | Mid | [
0.615598885793871,
27.625,
17.25
] |
Answering Your Questions: Mascara for Sensitive Eyes After my last mascara blog went up here and on Allure's Facebook page (become a fan there if you haven't already!), I noticed that a bunch of FB commenters asked about formulas that don't leave sensitive eyes red and watery. Good question! I asked makeup artist Troy Surratt for advice. One of the most important things to remember if you have very sensitive eyes, he says, "is to use a hypoallergenic or very basic formula like Shu Uemura Basic Mascara or Maybelline Great Lash." Surratt also recommends the ultra-sensitive stay away from: lengthening and thickening mascaras, "since they may contain fibers that could flake off as the day progresses and leave particles in your eyes, causing discomfort and sensitivity"; waterproof formulas, which contain added polymers and ingredients, and "more ingredients increase the risk of irritation"; and colored mascaras (especially purples, red-based browns, and burgundies—makeup-related eye sensitivities often are due to the use of red pigments). And we know you've heard this before, but it bears repeating: Throw out mascaras after three months of use. They begin to dry out after they're opened, and, Surratt says, "the drier the formula becomes, the more likely it is to break down on the lashes and flake into the eyes." PHOTO: COURTESY OF SHU UEMURA | Mid | [
0.6243523316062171,
30.125,
18.125
] |
Introduction {#sec1-1} ============ The recent World Health Organization (WHO) report and other studies suggest dental caries to be a major public health problem in most of the developing countries, affecting 60-90% of the school children in spite of the declining trends in the most developed countries.\[[@ref1][@ref2]\] The decline in dental caries among children in highly developed countries started to emerge around 1970 and the percentages of caries free children in different age categories have increased since then. This was mainly attributed to the increased use of fluorides from all sources, especially toothpastes.\[[@ref3][@ref4]\] On the other hand, certain developing countries, have reported an increase in dental caries\[[@ref1][@ref4]\] The economic, social and political changes in the developing world have had a significant impact on diet and nutrition, with a shift from traditional to a more westernized life-style. This is attributed to increased availability and consumption of refined sugars. There are few resources for curative/restorative intervention and no infrastructure upon which to base large-scale prevention measures. Dental caries is, therefore potentially of major public health significance in developing countries, and the need to focus on its prevention is a matter of urgency.\[[@ref4][@ref5]\] India is one among the 25 nations around the globe, showing increasing trends in dental caries and it exists as a smoldering disease that has ingressed its tentacles deep into those regions where there are inadequate resources of dental treatment, lack of public health awareness and motivation and increase in the utilization of refined carbohydrates.\[[@ref6]\] National oral health survey and fluoride mapping in India,\[[@ref7]\] found the prevalence of dental caries among 12 and 15-year-old children to be 53.8% and 63.1% respectively. A recent study among school going children in an endemic fluoride area in Andhra Pradesh found the prevalence of dental caries among 12 and 15-year-old children to be 55.3% and 57.3%. Prevalence of dental fluorosis was 73% among 12 years and 70.1% among 15 years children.\[[@ref8]\] Dental caries and dental fluorosis are two dental diseases among the school children in endemic fluoride areas. Around 17 states in India are endemic to dental fluorosis.\[[@ref9]\] Many laboratory, clinical and dental public health researchers after more than 70 years of research have concluded that, fluoride is a double--edged weapon, where its deficiency increases the risk for dental caries and excess consumption in the first 5-10 years of life increases the risk for dental fluorosis.\[[@ref10]\] The fluoride produces a dose-dependent effect on the dentition. However, this is not confined to increased caries resistance. Fluoride causes various disorders, together called as fluorosis, if accumulated above certain levels in the body. According to a WHO report "It may not be possible to achieve effective fluoride based caries prevention without some degree of dental fluorosis." Public health administrators must seek to maximize caries reduction, whereas minimizing dental fluorosis.\[[@ref11]\] The literature on the relation between fluoride concentration in drinking water with dental caries is conflicting. Some studies reveal an inverse relation\[[@ref12][@ref13]\] while others found no relation\[[@ref14]\] or a positive association.\[[@ref15][@ref16]\] Nalgonda District in Andhra Pradesh is an endemic fluoride area with fluoride concentration in drinking water ranging from below optimum to optimum and above optimum levels. The district has 1155 revenue villages with 3,359 habitations out of which, 1,122 habitations were identified as fluoride affected.\[[@ref17]\] The sparse literature on the relation between dental caries, dental fluorosis with fluoride concentration in drinking water among the school going children in Nalgonda district prompted us for the present study to assess dental caries and dental fluorosis prevalence among 12 and 15-years-old school children in Nalgonda district, Andhra Pradesh, India. Subjects and Methods {#sec1-2} ==================== This cross-sectional study was conducted among 12 and 15-year-old school children in Nalgonda district, Andhra Pradesh, India. The study was conducted over a period of 8 months from July 2009 to February 2010. Before the start of the main study, a pilot study was conducted on a convenient sample (*n* = 100) of school children. The pilot study found the prevalence of dental caries to be 30%. Based on this, the desired sample size was estimated to be 1613 with 95% confidence level and a design effect of five using Master software (Biostatistics Resource and Training Center, Christian Medical College, Vellore). Two stage cluster sampling was used for selection of study participants to the study. Nalgonda district has 59 mandals (administrative divisions) and the district was divided into four zones (North, South, East and West). All the mandals in these zones were listed and five mandals from each of these zones were selected in the first stage. All the government higher secondary schools in these selected mandals were listed. Using this as the sampling frame, one school from each of these 20 mandals was selected using lottery method of simple random sampling. The selected schools were listed and permission to conduct the study in these schools was obtained from the District Education Officer, Nalgonda and concerned head masters. The schedule for conducting the study among these children was prepared and sent to the concerned head masters well in advance. On the day of clinical examination, the list of all grade VI and IX children was procured and the details related to their date of birth were checked. All children from grade VI and IX fulfilling the following inclusion and exclusion criteria were examined in the study. Inclusion criteria {#sec2-1} ------------------ Children aged 12 years and 15 years (those who completed 12 years and 15 years on the day of examination)School children who were lifelong residents in that region and who were using one source of drinking water from birth to at least 10 years of their lifeChildren with all permanent teeth, except third molars, with at least more than 50% of the crown erupted. Exclusion criteria {#sec2-2} ------------------ Children aged less or more than 12 years as well as those aged more or less than 15 yearsMigrated children from some other place or who were not the permanent residents of the area concerned.Children with a history of drinking water from more than one source in the initial 10 years of their lifeChildren with orthodontic brackets were excluded as this hindered diagnosis of enamel defectsChildren with severe extrinsic stains on their teeth in whom assessing fluorosis was not possible. Ethical clearance for the study was obtained from the Institutional ethics committee, Sri Sai College of Dental Surgery, Vikarabad. A pre-designed structured questionnaire was used to collect the desired information such as oral hygiene practices, diet, sugar exposure, source of drinking water etc., The questionnaire had 15 close ended questions with multiple options to collect information on these aspects. Questionnaire was filled by the investigator by means of face to face interview to avoid misinterpretation of questions and to ensure uniformity in data collection. One trained and calibrated examiner conducted the clinical examination of the selected children using a mouth mirror and community periodontal index probe. The Kappa co-efficient value for intra-examiner reliability was found to be 0.88 for decayed missing filled teeth (DMFT) and 0.81 for dental fluorosis. The agreement was substantial to almost perfect, according to the scale of Landis and Koch.\[[@ref18]\] The clinical examination was carried in the school premises under natural day light on the plastic chair after obtaining informed consent from the children and their parents. Participants were made to sit on a chair in an upright position using wall as the head rest, whenever necessary. Dental caries was assessed using Dentition status and treatment needs and dental fluorosis using Dental fluorosis index (WHO oral health surveys).\[[@ref19]\] The first child in the school was requested to obtain 500 ml of water from the source from which children consumed water. The investigator/assistant accompanied the child at the time of collection of water sample. The other samples in the school were collected only when the source of drinking water differed among the children. All the water bottles collected from each school were coded and sent to laboratory for estimation of fluoride concentration. The code written on the water bottle was entered on all the data collection sheets of the children who consumed water from the respective source. This ensured the collection of water samples from all the sources from where the eligible children consumed water during their childhood. The estimation of fluoride concentration in the drinking water was done at "National Institute of Nutrition (Indian Council of Medical Research) Hyderabad using Orion 720 A fluoride meter, coupled with ion specific electrode. The data was entered onto a personal computer and statistical analysis was done using the statistical package for the social sciences. Inc. version 16, (Chicago, USA). The children were classified into four categories based on the fluoride concentration in the drinking water. This was done at the time of statistical analysis. This ensured that the investigator was not aware about the fluoride concentration in the area at the time when the clinical examination of the children was performed. The four categories were low fluoride area (fluoride concentration \<0.7 ppm), medium (0.7-1.2 ppm), high (1.2-4 ppm) and very high fluoride area (4-6.28 ppm). Mean and standard deviation were used to express the quantitative data, while the qualitative data was presented in frequencies and percentages. Kruskal walley\'s test was used to compare the difference in the mean DMFT between different categories. The prevalence of dental caries and dental fluorosis was compared using Pearson\'s Chi-square test. Spearman\'s rho was used to correlate fluoride concentration with dental caries and dental fluorosis. The statistical significance was fixed at 0.05. The autoclaved set of instruments was used for oral examination of the children. The methodology employed in the study is diagrammatically depicted in [Annexure 1](#App1){ref-type="app"}. Results {#sec1-3} ======= A total of 1875 school children were examined in the present study. The gender distribution of the study population in different fluoride areas is denoted in [Table 1](#T1){ref-type="table"}. ###### Age and gender distribution of the study population in different fluoride areas  Dental caries prevalence {#sec2-3} ------------------------ The overall prevalence of dental caries among the school children was 43.4% (813/1875). The prevalence of dental caries was significantly higher among females (50.4% \[492/977\] compared with males (35.8% \[321/898\]). The prevalence was more among 15-years-old children (46.7% \[444/951\] compared with 12 years children (39.9% \[369/924\]). The prevalence of dental caries among children in low fluoride areas was 60.5% \[300/496\] followed by very high fluoride area (54.8% \[201/367\]), high fluoride area (32.4% \[293/904\]) and medium fluoride area (17.6% \[19/108\]). The difference in the prevalence of dental caries between different fluoride areas was statistically significant. These results were true even when the comparison was made separately among 12 and 15 years old children \[[Table 2](#T2){ref-type="table"}\]. ###### Prevalence of dental caries among 12 and 15 years old school children in different fluoride areas  Dental fluorosis prevalence {#sec2-4} --------------------------- The prevalence of dental fluorosis was 76.8%. There was no statistically significant difference in the prevalence of dental fluorosis between 12 and 15-years-old children. There were no gender differences in the prevalence of dental fluorosis. The prevalence of dental fluorosis significantly increased with increasing fluoride concentration. The prevalence of dental fluorosis in very high, high, medium and low fluoride areas was 100% (367/367), 96.6% (873/904), 47.2% (51/108) and 29.8% (148/496) respectively. These results were true even when a separate comparison was made between different fluoride areas among 12 and 15-year-old children \[[Table 3](#T3){ref-type="table"}\]. ###### Prevalence of dental fluorosis among 12 and 15 years old school children in different fluoride areas  Correlation between fluoride concentrations in water and dental caries, dental fluorosis {#sec2-5} ---------------------------------------------------------------------------------------- The mean DMFT score was lowest in the medium fluoride area (0.22 \[0.49\]). The mean DMFT score in high fluoride area, very high fluoride area and low fluoride area was 0.46 (0.82), 0.92 (1.02) and 1.18 (1.13) respectively. The difference in the mean DMFT score between different fluoride areas was statistically significant \[*P* \< 0.01, [Table 4](#T4){ref-type="table"}\]. A negative correlation (Spearman\'s *P* = −0.18) was noted between fluoride concentration in drinking water and mean DMFT score. A positive correlation (Spearman\'s *P* = 0.92) was observed between dental fluorosis index score and fluoride concentration. There was no statistically significant difference in relation to dietary habits, oral hygiene aids used, frequency of brushing and frequency of sugar consumption among the school children in different fluoride areas. ###### Correlation between fluoride concentration in drinking water with mean DMFT and fluorosis  Discussion {#sec1-4} ========== The scanty published literature on the prevalence of dental caries and fluorosis among school children in endemic fluoride areas, which provide the best opportunity in assessing the relation between increasing fluoride concentration on both these diseases prompted us to undertake the present study. The study was conducted among 12 and 15-year-old school children in Nalgonda district, Andhra Pradesh, India. The WHO in its manual on basic oral surveys (1994)\[[@ref19]\] has specified 5, 12, 15, 35-44 and 65-74 years as index age groups for assessing the oral health status. Among these, 12 years is considered the global monitoring age for international comparison of dental caries and 15 years represent the other adolescent age group. Moreover, these are the two age groups for obtaining a reliable sample from the school system. This prompted us in the selection of 12 and 15-years-old school children as the study participants in the present study. The prevalence of dental caries was least in medium fluoride area followed by high fluoride area. The prevalence was highest in low fluoride area followed by very high fluoride areas. The mean DMFT value was also least in medium fluoride area. The highest mean DMFT was noted among children in low fluoride area followed by very high fluoride area and high fluoride area in the descending order. This clearly suggests that fluoride in drinking water offer maximum protection against dental caries at optimal concentration which ranges from 0.7 to 1.2 parts per million (ppm) according to previously published literature.\[[@ref8]\] The fluoride concentration below optimal as well as above optimal was found to be harmful. A study by Shekar *et al*.\[[@ref8]\] among 12 and 15-years-old children in Nalgonda district, Andhra Pradesh found the prevalence of dental caries to be highest in areas where the concentration of fluoride in drinking water was less than 0.7 ppm followed by areas with fluoride concentration of 4.1 ppm and above. The results of the present study were consistent with the findings of this study. Murray JJ (2003)\[[@ref20]\] quoted a study by Torell and Ericsson assessing the use of fluorides in different forms for caries prevention. The study found the frequent exposure to low concentration fluoride solution to be more protective than infrequent exposure to high concentration fluorides. The continuous exposure of the teeth to water containing optimum and slightly above optimum levels of fluoride may facilitate remineralization of incipient carious lesions through topical effect. The optimum fluoride concentration though may result in milder forms of dental fluorosis will not alter the surface texture of the teeth at these concentrations. The fluoride concentrations above 4 ppm may result in severe dental fluorosis which manifest as confluent pitting.\[[@ref20]\] The alteration in the surface texture favors food retention and plaque accumulation, which in turn may increase the risk for dental caries in areas where the concentration of fluoride is above 4 ppm. The lack of protective benefit and pitting caused by severe dental fluorosis explain the high caries prevalence in low and very high fluoride areas respectively. Reddy and Tewari.\[[@ref21]\] conducted a study among 1750 school children in the age group 12-17 years in Bhatinda district of Punjab, India. The prevalence of dental caries in areas with the fluoride concentrations of 0.3, 1.1, 2, 3.4, 5.4 and 10.4 ppm was 89.0, 61.2, 54.7, 72.8 73.6 and 85.5% respectively. The study found highest caries prevalence in low fluoride areas (0.3 ppm) and very high fluoride areas (10.4 ppm). The results of our study were consistent with the findings of this study and others.\[[@ref15][@ref22][@ref23][@ref24][@ref25][@ref26][@ref27][@ref28]\] We found a higher prevalence of dental caries among females compared with males. In the absence of dietary differences, the higher caries prevalence among females is attributed to early eruption of teeth among females and thereby longer exposure to deleterious oral environment.\[[@ref29]\] The body surface area among males is more than that for females. Due to greater physical activity, boys consume more water than compared to girls. The frequent water consumption results in frequent exposure of teeth to fluoridated water and higher level of protection among boys compared to girls. A study by Singh and Singh\[[@ref30]\] among school going children in Patna found the prevalence to be high in female children (52.6%) than males (50.2%). The results of our study were in agreement with this study and others.\[[@ref31]\] The irreversible nature of the disease and longer exposure of teeth to deleterious oral environment may explain the higher caries prevalence of caries among 15 years group compared to 12 years group. National oral health survey and fluoride mapping in India\[[@ref7]\] found the overall prevalence of dental caries to be increasing with increasing age. Our results were consistent with the results of national oral health survey in India. The overall prevalence of dental caries among 12 and 15-year-old children in the present study was significantly less compared to prevalence noted in National oral health survey in India for these age groups. Our study was conducted in an endemic fluoride belt and hence, the less caries experience among children in our study compared to data at national level may be attributed to protective action of fluoride in drinking water. The increasing prevalence and severity of dental fluorosis with increasing fluoride concentration may be explained by the fact that dental fluorosis is a developmental defect which occur because of exposure to water containing high fluoride concentrations. This relation between water fluoride concentration and severity of dental fluorosis is dose dependent with increasing concentration leading to higher risk.\[[@ref32]\] A positive correlation between fluoride concentration and dental fluorosis indexscore was found in many epidemiological studies in the past.\[[@ref25][@ref33][@ref34]\] our results were consistent with all these studies. Dental fluorosis is a developmental defect affecting the teeth before calcification (before 10 years of age). The exposure to higher fluoride concentrations after the calcification (which is complete by 10 years of age for all permanent teeth except third molars) might not increase the severity of dental fluorosis. This explains the lack of age and gender predilection for dental fluorosis in the present study. Chandrashekar and Anuradha\[[@ref35]\] in their study on the prevalence of dental fluorosis in rural areas of Davangere, India found no age predilection for dental fluorosis. Our results were in agreement with this study and others.\[[@ref36]\] The number of children in medium fluoride areas in the present study was significantly less compared to other categories. The fluoride categories were made at the time of statistical analysis based on fluoride concentration in drinking water and not at the time of selection of study participants. This clearly suggests that the optimal fluoride areas are less in number even in an endemic fluoride belt. Many areas were having either below optimal fluoride concentration or above optimal levels. This might have resulted in less number of children in medium fluoride areas. The finding highlights the need for a comprehensive fluoride mapping in endemic fluoride belts and remedial actions based on fluoride concentrations in water. The less number of children in medium fluoride area compared to others might have influenced the difference between different fluoride areas to some extent, which needs to be further evaluated in future studies. Summary and Conclusion {#sec1-5} ====================== The present study found high caries prevalence in low and high fluoride areas with fluoride concentrations lesser than 0.7 ppm and higher than 4 ppm. The medium fluoride area with optimum fluoride concentration (0.7-1.2 ppm) had the lowest caries prevalence. The prevalence of caries was more among females compared to males. 15-years-old children had more caries prevalence than 12 years. There was a positive correlation between fluoride concentration and prevalence of dental fluorosis with no age and gender predilection. The results strongly support the findings of previous studies conducted in endemic fluoride areas that increase in the fluoride concentration above optimal levels does not offer additional benefits in caries prevention. But, instead, increase the risk for dental fluorosis. A defluoridation plant was installed in Nalgonda by National Environmental Engineering and Research Institute, Nagpur. This community defluoridation technique is popularly known among dental fraternity in India as Nalgonda technique of defluoridation. The defluoridation plant is presently abandoned due to lack of maintenance. This reflects the negligent attitude among the policy makers towards dental diseases. The results of this study and previously conducted studies highlight the immediate need for defluoridation at least in areas with fluoride concentration above 2 ppm. Here, dental fluorosis is a major public health problem than dental caries as the prevalence is almost 100%. The high caries prevalence in low fluoride areas call for water fluoridation and or other alternate strategies for caries prevention. The running and maintenance of defluoridation plants is more expensive than the fluoridation plants. Determined efforts with public private partnership can translate these idealistic goals into realistic ones. The professional organizations such as Dental Council of India and Indian Dental Association should strongly support water fluoridation in low fluoride areas and defluoridation in high fluoride areas to combat two dental diseases related to fluoride concentrations in drinking water. I would like to thank the Principal and management of Sri Sai College of Dental Surgery for their continuous support and encouragement for undertaking this project. We thank the District Education Officer, Nalgonda, The Head masters of the schools concerned and the participants for their kind co-operation. **Source of Support:** Nil. **Conflict of Interest:** None declared.  | High | [
0.6596306068601581,
31.25,
16.125
] |
1. Introduction {#sec1} =============== Acute fatty liver of pregnancy (AFLP) is an uncommon, life-threatening complication that occurs in the third trimester or early postpartum period \[[@B1]\]. In the literature, only few cases of acute fatty liver of pregnancy in the second trimester were reported \[[@B2], [@B3]\]. Here, we report a case of AFLP with a milder clinical presentation in the second trimester. 2. Case Report {#sec2} ============== A 29-year-old woman, gravid 1, presented at 23 weeks\' gestation with nausea, vomiting, abdominal pain, malaise, and weight loss. An ultrasound at 11 weeks\' gestation determined gestational age. She was normotensive with a normal platelet count and no proteinuria. She had elevated liver transaminase levels (aspartate aminotransferase \[ASAT\] level 108 U/L, alanine aminotransferase \[ALAT\] level 104 U/L), prothrombin time \[PT\] 96%, negative hepatitis serology (A, B, C, and E), and a negative immunologic analysis. She was treated with intravenous fluids, but we remarked a slightly aggravation of her laboratory findings ASAT 122 U/L, ALAT 248 U/L, GGT 18 U/L, PAL 118 U/L, bilirubin total/direct 23,2/17,4 *μ*mol/L, PT 76%, WBC 7800, platelets 178000, glucose 65 mg/dL, and uric acid 246 *μ*mol/L. The abdominal ultrasound showed a brilliant liver without others anomalies. Because of the doubt on diagnosis and stable clinical and biologic parameters, a liver biopsy was performed (10 days after admission) and showed a microvesicular fatty infiltration of the hepatocytes ([Figure 1](#fig1){ref-type="fig"}). A diagnosis of acute fatty liver of pregnancy was confirmed. The decision was to prolong the pregnancy under surveillance in obstetric unit care with clinical and biological control. The biological analyses were stable ([Table 1](#tab1){ref-type="table"}) with good clinical condition. The evolution was good and the pregnancy was prolonged until 36 weeks of gestation, and the patient gave birth to a healthy newborn of 2900 g, Apgar score of 9/10. The delivery was simple without any complications. After delivery, all biologic parameters were normalized within 7--10 days and the patient was discharged. 3. Comments {#sec3} =========== Acute fatty liver of pregnancy is a rare and potentially fatal condition. It has been thought that this condition is rare and usually takes a devastating course with significant morbidity and mortality \[[@B1]\], incidence is estimated to be 1/1000 to 1/20000 deliveries \[[@B1]\], and, generally, it occurs in the third trimester or early postpartum period of pregnancy, but a few cases in the second trimester were published \[[@B2], [@B3]\]. Furthermore, modern advances in diagnostic capabilities have given us a better understanding of the pathological process, and many cases with a milder clinical course as well as subclinical cases have been reported \[[@B4]--[@B6]\]. Our observation is unusual by its clinical evolution until term and time of arising (second trimester). Indeed, Monga and katz \[[@B2]\] reported a case of AFLP arising at 22 weeks\' gestation in 35 years women. However, in this case evolution was interrupted because of a disseminated intravascular coagulation. Also, Suzuki et al. \[[@B3]\] reported a case arising at 23 weeks of gestation for which the interval between symptom occurrence and fetal demise was only one week and pregnancy was quickly interrupted. Curiously, biological findings in our observation were constant during evolution of pregnancy, particularly, prothrombin test and platelet counts. These findings may confirm the cases with subclinical or milder clinical course already reported. Usually, there is no specific treatment for acute fatty liver of pregnancy except prompt termination of pregnancy and pathological changes as a rule improve rapidly after delivery. Liver biopsy is the best way to confirm the diagnosis, but because it is invasive, it is not always performed. Furthermore, today we can take advantage of noninvasive procedures to demonstrate fat in the liver and to exclude other liver diseases such as viral hepatitis \[[@B7]\]. Histopathological finding in liver biopsies is small and midsized intracellular vacuoles, without displacement of the nucleus and ballooning hepatocytes. Recently, an association between AFLP and a deficiency of the enzyme long-chain 3-hydroxyacyl-CoA dehydrogenase was suggested. This enzyme is one of the four enzymes, which break down long-chain fatty acids in the liver. Deficiency of this enzyme results in the increased accumulation of long-chain fatty acids \[[@B8]\]. However, the exact mechanism that explains these mild clinical presentations remains unknown. This case, one of few reported cases, shows that even in the second trimester, the diagnosis of acute fatty liver should be considered when a woman presents with deteriorating liver function, jaundice, and coagulopathy. The particularity of this case is the unusual evolution; with a stable clinical and biological perturbations for 13 weeks after diagnosis of AFLP. Finally, the diagnosis of AFLP must be suspected in front of all digestive symptomatology in the second or the third trimester of pregnancy. {#fig1} ###### Biologic analysis evolution. Day after admission 1 3 7 21 42 63 84 91 --------------------- -------- -------- -------- -------- -------- -------- -------- -------- ALAT IU/L 104 248 213 165 172 146 134 145 ASAT IU/L 108 122 104 106 110 96 98 102 PAL IU/L 138 118 123 121 142 134 129 119 BIL T *μ*mol/L 14 23,2 18,4 16,7 14,5 16 15,8 17,3 BIL D *μ*mol/L 8,5 17,4 12,6 11,2 10,4 12,1 11,3 12,9 Creatinine *μ*mol/L 64 78 67 56 64 61 73 57 Uric acid mmol/L 144 138 145 164 154 151 167 173 Hemoglobin g/dL 12,2 12,3 11,9 12 11,5 11,7 11,4 11,8 Platelets/mm^3^ 213000 178000 198000 187000 201000 221000 197000 184000 WBC 7600 7800 8500 9200 7400 10100 10700 12000 PT % 96% 76% 78% 81% 92% 84% 79% 85% Glucose mmol/L 3,6 4 4,9 5,1 4,4 5,3 4,8 5,5 WBC: white blood cell. PT: prothrombin time. ASAT: aspartate aminotransferase. ALAT: alanine aminotransferase. [^1]: Academic Editor: E. Cosmi | High | [
0.670212765957446,
31.5,
15.5
] |
I ordered my first Kimber .45 and cant wait to shoot it. Its the Kimber Stainless Pro TLE ii with LG grips. I just had a few questions to put out there about the gun. Does anyone own this gun? How is it?... reviews, pros/cons., etc.? And also what is a good break-in ammo and a good home defense ammo? Any responses will be helpful. Thanks 04-13-2010, 05:31 PM Allterrain Well, first of all I think I'm going to have to stop reading posts on this forum. Everybody talks about there new gun and it just makes me want another one!!!! I'm gonna go broke on this site!!!. Congrats on your Kimber. Dont own one but I can tell you what kind of ammo I use for self defence, and without a douht I use and fully trust the remington golden sabre bullet. I keep a mag filled with 9mm and 45. Use the cheep stuff to target practice. 04-15-2010, 05:38 PM cougartex Congratulations. For target practice use any brand name FMJ, I agree about the Remington Golden Saber for HD. | Mid | [
0.589430894308943,
36.25,
25.25
] |
JCppEdit Software JCppEdit is an integrated development environment (IDE) solution that allows programmers to write code in multiple programming languages using a single platform. The solution is deployed on-premise and works on Windows 7,8 and 10. JCppEdit supports programming languages including C, C++, HTML, CSS, JS, XML and JAVA, and also helps users identify programming errors in real time. The platform offers an autocomplete tool that gives suggestions to programmers while typing. An autosave feature can also be enabled. | High | [
0.66824644549763,
35.25,
17.5
] |
{% extends 'LadbCoreBundle:Common:Modal/_modal.part.html.twig' %} {% set id = 'edit_'~formSection~'_'~picture.id~'_modal' %} {% set title = 'picture.edit'|trans() %} {% set bodyClasses = 'row' %} {% form_theme form 'LadbCoreBundle:Common:_form-theme.twig.twig' %} {% block contentInner %} {{ form_start(form, { 'action':path('core_picture_update', { 'id':picture.id, 'formSection':formSection, 'sortable':sortable }), 'method':'POST' }) }} {{ parent() }} {{ form_end(form) }} <script> $("#{{ id }}").find('form').ajaxForm({ cache: false, dataType: "html", context: document.body, success: function(data, textStatus, jqXHR) { if ($(data).find(".alert-danger").length == 0) { $("#{{ id }}").modal("hide"); $("#{{ id }}").remove(); $("#picture_{{ picture.id }}").replaceWith(data); } else { $("#{{ id }} fieldset").parent().replaceWith(data); } $("#{{ formSection }}_{{ picture.id }}_save_btn").button('reset'); }, error: function() { $("#{{ id }}").modal("hide"); $("#{{ id }}").remove(); $("#{{ formSection }}_{{ picture.id }}_save_btn").button('reset'); } }); setupTooltips(); </script> {% endblock %} {% block bodyInner %} <div class="col-xs-12 ladb-margin-bottom text-center"> <div style="width: 100%; height: 300px; line-height: 300px;"> <img id="{{ id }}_preview" src="{{ ('/'~picture.masterPath)|imagine_filter('300x300i') }}" class="thumbnail ladb-no-transition ladb-rotate{{ form.rotation.vars.data }}" style="display:inline-block; vertical-align: middle; line-height: normal;"> </div> <div class="text-center ladb-margin-top"> <a href="#" class="btn btn-default" onclick="pictureGalleryRotatePreview($('#{{ id }}_preview'), $('#{{ form.rotation.vars.id }}'), -90); return false;" data-tooltip="tooltip" title="{{ 'picture.turn_left'|trans() }}"><i class="ladb-icon-rotate-left ladb-icon-lg"></i></a> <a href="#" class="btn btn-default" onclick="pictureGalleryRotatePreview($('#{{ id }}_preview'), $('#{{ form.rotation.vars.id }}'), 90); return false;" data-tooltip="tooltip" title="{{ 'picture.turn_left'|trans() }}"><i class="ladb-icon-rotate-right ladb-icon-lg"></i></a> </div> </div> {% include 'LadbCoreBundle:Core/Picture:_fieldset.part.html.twig' %} {% endblock %} {% block footerInner %} <a href="#" class="btn btn-default" onclick="pictureGalleryCancelEditPicture('{{ formSection }}', '{{ picture.id }}'); return false;">{{ 'default.cancel'|trans() }}</a> {% include 'LadbCoreBundle:Common:_form-btn-loading.part.html.twig' with { 'id':formSection~'_'~picture.id~'_save_btn', 'text':('default.save'|trans()), 'classes':'btn-primary' } %} {% endblock %} | Low | [
0.5083135391923991,
26.75,
25.875
] |
The warning phase of the ‘Extreme Heat and Ozone Peaks Plan’ was activated on Friday after the Royal Meteorological Institute, IRM, predicted high temperatures throughout Belgium. Temperatures are expected to rise over the next few days, climbing close to 35 °C on Tuesday next week, according to IRM. Ozone concentrations in the air are also expected to increase, although it is still too early to say whether the European information threshold will be exceeded. The warning phase of Belgium’s Extreme Heat and Ozone Peaks Plan is automatically activated when a certain cumulative threshold of heat is exceeded in the IRM’s five-day weather forecasts. The activation is based on the observation of highs expected over the next five days. The main aim of the warning phase is to inform the public, as well as health professional and personal-care staff to be particularly watchful and to make sure they take precautions such as checking the reserves of liquids and temperatures inside buildings. The public is also advised to pay close attention to loved ones and isolated persons, especially if they belong to high-risk groups such as the elderly, the chronically ill, babies and children. Precautions advised in the event of extreme heat or ozone peaks include drinking enough mineral-rich water, staying indoors as much as possible (during ozone peaks) and avoiding alcohol and sweetened beverages. Oscar Schneider The Brussels Times | High | [
0.6666666666666661,
37.75,
18.875
] |
Node-skeleton, a quick start to Microservices - tvvignesh https://github.com/tvvignesh/node-skeleton ====== tvvignesh Hi. Vignesh here - Author of Node-skeleton You can find the article I wrote on the same here: [https://medium.com/techahoy/building-a-boilerplate-for- micro...](https://medium.com/techahoy/building-a-boilerplate-for- microservices-part-1-166ce00f5ce9) | Mid | [
0.620155038759689,
30,
18.375
] |
"I am a privacy fence salesman." "Top sales, three straight years." "Goal -- make manager by next spring." "Quit selling dead people's crap on the weekends." "Get excited about my wedding." "Hey, man." "Want to do a couple funerals?" "'Sup, douche?" "Go time, playa." ""You die, and we fly."" "Know what I'm sayin'?" "Know what I'm sayin', Trevor G?" "William Zill." "Centerville." "Sounds like a loser." "Yeah, that one looked hot." "Connie Campbell." "Remember her three years ago?" "Yeah." "The unicorn collection got me my flat-screen." "Two weeks till they sell the house." "Mm." "Let's get the shit, hombre." "This is the place we've been looking for." "Hey." "I saw on the Internet that shrimp fettuccine is only 293 people in the whole world's favorite food -- out of 2 million." "I'm special." "Speaking of special -- did you and the little lady set a date yet?" "No, Rob." "Still waiting for the... .. .pagoda." "What's a pagoda?" "Ay, caramba." ""F."" "This looks like one of my mom's Bible-study buddies' places." "Ew!" "You're old enough for your own condo, man, especially with the big promotion." "Promotion." "F'ing l.T. Wage wookie man." "I swear, every one of these places is getting worse." "Bunch of junk." "Look at that." "TV ." "Yep." "All this crap from Bang a Gong ain't selling on eBay, man." "Word." "We'll see what he's got upstairs." "Bunch of junk." "Yeah, baby." "Whoops." "Nice porn." "What do you know?" "Another nut." "What do you know?" "Why everybody got ta be so weird?" " Computer." "Yep." "Oh, shit." "Trevor, check it out." "This guy's seen "Four Boxes."" "What's "Four Boxes"?" "I'll show it to you." "Seriously, I thought I was the only one that knew about it." "It's this freaky little deep Google site I found googling cheap cardboard boxes." "Fourboxes.tv." "Regular voyeur site, whatever." "Chick put tiny, invisible cameras up all over her apartment." "Watch this chick eat cereal, flick the bean, get ready for work, blah, blah, blah." "But, no, then it starts to rock." "Couple weeks ago, chick moves out, takes her shit, but didn't unhook the cameras." "Then, a couple days ago, guy rents the place -- fourboxes.tv -- to a totally new person." "There he is." "This guy doesn't know he's being watched." "And the dude trumps the chick." "Man, he's fing weird." "I call him Havoc." "Where is this coming from?" "No way to know, man." "Four rooms, four boxes full of plastic crap from China." "I mean, where's it not coming from, yo?" "Sound?" "Silent movie." "Doesn't look like a chick's place." "l know." "Where's the Glade Plugins?" "It did, man." "A hot one, too." "Bummed to see her go." "She had an awesome ass." "This guy brought all this crap in and totally just trashed the place." "Hey." "What's up with this?" "Oh, yeah." "Cops said this dude's ex-wife hung herself over here six months ago just to fuck with the guy." "Suicides rule!" "Like my dad -- get out of our way, losers." "I love it, man." "I love it." "When my parents were still married, this is all they had..." "everywhere." "I'm probably gonna have to waste some money and pick up that new Chili Peppers disc when I get a sec." "It's out today." "Anthony Kiedis." "Flea." "It sucks when suddenly you realize you're old school." "Yeah, it sucks." "Play some live, for Christ sakes." "Life sucks, dude, right?" "Yep." "Life sucks." "# Ooh #" "# Ooh #" "Shower." "Yeah." "No fing way!" "Yeah, baby." "Losers rule!" "What is this shit?" "That's about 500 hollow eggs." "Did you see the photo album of this dude's upstairs?" "All the faces are cut out of the pictures." "BFD." "My mom cut Tim's face out of all my baby pictures." "Yeah, one big difference here, homeslice -- these pictures ain't got no fat, ugly bitch in them next to the hole." "True dat." "That is pretty weird, I guess." "It's like someone was trying to get rid of themselves." "Man..." "I remember when I was a kid just laying on my bed listening to the fucking shit out of "Californication."" "Wasn't that just like five years ago?" "Yeah, I guess." "I don't know." "Oh." "Check these, man." "Mm." "Get in that basket, bitch." "What's she here, dude?" "Chips-and-dip bowl, I guess." "Computers make me horny." "One cool thing the Internet makes you realize -- the only difference of all the world's chicks -- l mean, with their clothes off -- is their teeth and the fucking couches they're spread-eagle on." "If it's like a burlap bag for a couch, then it's probably, like, Russia." "Word." "Should we just sleep here, then?" "Why not?" "Easy." "Mmm. I got to try to start drinking wine now." "How come?" "I've got to grow up." "Wack place." "Hey, babe." "Hey, honey bundle." "And hi, Trevor." "Got some breakfasts." " Sweet." "I'm flippin' ravenous over here." "What are you doing here, Amber?" "Hey, I invited her, man." "What?" "Why?" "Um, because she's my fiancee, and I wanted her here." "What about our work, homes?" "It's my choice, bro, and I wanted her here." "Any of you guys remember last summer?" "I introduced you guys at TGl Bardays?" "You, my girlfriend?" "You, my junior partner?" "You did me dirty, man." "Trevor, things just happen." "Fate and chance and" "Oh, here we go." "It's, like, don't you got a new girlfriend now, too?" "Things flow, Trevor." "Things are." "Every day's a new world." "Yeah." "Brit." "We're engaged, actually." "Brit." "That's right " " Brit." "Well, see, that's grand." "That's grand." "You sound like your mom." "Really?" "Well, awesome." "She's my best friend, so" "That bitch is your best friend?" "Not cool." "Hey, man." "Come on." "You might have fucking told me the plan, partner." "Yeah, but it was my plan, broseph." "Plus, I knew if I told you, you would never have gone for it." "Oh, flubber!" "What happened, cooch?" "Oh, no biggie bigsville." "Just munched a $45 manicure in the ass." "Girl thing." "Still really ticked about my futon, too." "Futon?" "What futon?" "Yeah." "Yeah." "Yeah." "This is super awesome." "Yeah." "Do it righteous." "Yeah." "Totally awesome." "Yeah!" "Oh, keep having sex!" "I love it!" "This is so strange." "What's that?" "Whatever, Rob." "What are you listening to?" "Music?" "Yeah." "I had to drown out the racket." "Hmm." "Sunset." "Sunset." "Aha." "Ah, shit." "What the fuck?" "# All we ever talk about #" "# Is all the things we think about #" "Organize my garage." "# Like love and pain and loneliness #" "Dump Brit." "# And me and you and tenderness #" "# And now #" "# Now #" "# And now I'm on my own again #" "# And learning how to love again #" "# And learning how to love again #" "La choy is in the hizz." "Why am I hungry again?" "Rankrizzle." "You just pile-drived, pile-drove, like 10 egg rolls." "God." "This bed is like holy hell." "I knew this would be weird with Trevor." "Us planning our wedding, and him -- so out of touch with the flow of his emotions." "You totally still have a thing for him, don't you?" "Someone posted something on my site." "What, on "Tears of sadness"?" ""Tears of confusion."" "Jesus, Rob." "Learn my site name." "Oh." "A pervert in Colorado thinks I'm hot." "Awesome." "Yeah, it is awesome." "What I mean is maybe it wasn't copacetic for us to come up here, or down here or wherever we are, you know what I mean, and, like, have everything be all out of tune." "Rob, it's like... you know, it's tough news, but it's true." "Trevor and I used to rump-ride, and that's always a BD, a big deal, you know?" "I mean, come on." "He's, like, out of tune with his, like, emotions, you know?" "I mean, and I know that's not our problem, you know?" "It's, like, he's the one that needs to be dealing with it." "Like, we can't come up here, come down here" "Like, what the fuck are you even saying?" "I mean, seriously." "I can't even follow it, actually." "Whatever." "Whatever." "Jesus." "God, I look fat with my guitar." "Oh, shit." "I forgot to pick up the new Chili Peppers disc today." "Creepy." "If you watch too much Web, it makes you a loser." "You know that, right?" "Look!" "Oh, my God." "Those are feet." "That's so messed." "Remember, dudes, everything on the Internet is bullshit." "In fact, everything's bullshit everywhere." "Shouldn't we tell someone about that?" "Are you stoned?" "Yeah." "I'm always stoned, Trevor?" "Is that it?" "It helps me balance." "Actually, I'm thinking this is maybe Europe." "God!" "I should not have shown you that website, man." "I forgot how you get all..." "Ease up, man." "What'd I do?" "You stoned, too?" "Yeah." "I could not live in a dirty shithole like that." "I swear to God." "I second that emotion." "What you doing, bro?" "Dude, not everything is a Stephen King book, yo." "Are you having your period?" "My bad." "She's the one that got me riled." "What'd I do?" ""Thistle, your tits are hot, but your singing sucks."" "Like, who would take the time to sit down and type this?" "Dudes." " Oh, this guy." "Dudes, dudes, dudes, dudes!" "Oh." "Told you everything on the Web is bullshit." "I'm outie." "Amber!" "I'm gonna run and get a southwestern ranch hand on ciabatta." "You want one?" "Okay." "Awesome." "Where's the Subway?" "By BP." "Finally got a remote car starter, huh?" "Total lifesaver, Trevor." "BRB." "Southwestern rizzle hizz." "Sauce?" "Three cheddar churn?" "Hmm." "Nice." "Where's Rob?" "Sleeping." "Gotta hit the gym." "Gettin' fat." "Tell me about it, dawg." "What happened to your Saturn?" "Oh, I backed into a pole at work." "I screwed up the wheels but good." "Drunk?" "Mm, yeah." "I was celebrating." "I had a big gig at Tracey's store." "The candle and incense store at Tamarack?" "That was your big gig?" "Yeah, Trevor." "It was a great turnout." "Thought it closed." "This was before, a while ago." "Rob ever find out about the shitface shack?" "No." "And he better not." "I'm really with him now, Trevor." "I'm 30 years old." "I'm finally in a place where l got to think about my future." "l got to grow up." "l know." "I'm going to bed." "I get it." "I know." "Trevor, grown-ups do what they don't want to do." "I know!" "Awesome!" "Yes!" "Oh, God!" "Yes!" "Awesome!" "Dot." "Watching." "I'm dot." "House." "My." "Have you." "If." ""lf you have my money, you never leave this house alive." "I'm watching."" "What the fuck?" "Who wrote that?" "Oh, man." "Hey, Rob!" "God damn, dude!" "Smoking again?" "I've so got to quit." "Ew. I need some gum bad." "Check this." "Oh, my God." "What the F?" "And I think that leaf blower just blew the schnikes out of a kitty cat." "Okay, I'm sorry, but this site is getting really wack." "Mmm." "Are you still on Ambien?" "I have nightmares, Trevor." "Somebody should get all up in this asshole's grille." "You're right about that." "I miss my computer." "You could go home." "Mm, too scary there." "My roommate's in Tampa." "Lucky ducky." "I want to be in Tampa." "# Don't take away #" "# Another piece of me #" "# Don't break away #" "# Another part of me #" "# Don't leave me #" "# Crying to myself #" "# Don't make me #" "# Come crawling back to me #" "I can feel my hair getting all screwed up when I sing." "Where's Rob?" "Brown." "Did Bill Zill succumb?" "What?" "Did Bill Zill die?" "Yeah." "Trevor Grainger, I'm getting really tipsy." "Take me inside." "Trevor G, we are way out of line here." "I know, but I feel out of line." "But I want to start acting like I should, though." "Fuck!" "Oh, shit!" "Shit!" "Shit!" "Shit!" "What's up?" "Good morning, dollface." "Just grabbing -- grabbing a bloody for the shower." "Good choice." "You smell like Tommy for Men." "Dude, what are you even talking about?" "Looks like you two been eating the same Krispy Kreme at the same time." "Chill, Chachi." "You're dreaming." "Just talking, dude." "Don't forget, we used to do it, you know?" "We have stuff to talk about." "Yeah, I know." "You used to do it." "Everyone keeps reminding me." "Don't forget whose idea to have her come here, hombre." "She's my fiancee, man." "I don't usually say shit like this to you, Trevor, but get over it." "Me and her -- we're getting married, dude." "We're gonna push plastic strollers full of babies around Disneyland, a'ight?" "You're just trippin', man." "Yeah." "That's what I'm hoping." "Oh, there's your mystery disk computer." "Wake it up, man." ""Shooting Spree"?" "Yep." "Should we play a little before we ship the TV?" "That thing is fucking old school." "Hmm." "They're all upstairs." "They're all hiding in the bathroom, you fucking idiots!" "There was something strange going on in this house." "There's something strange going on inside every house, yo." "Eh!" "Suck it, mall walker!" "Rob!" "Come check this!" "Terrorists." "Yep." "And Havoc's the scientist." "It's just like in movies." "A terrorist hires a scientist to build him some bombs and then knocks him off when he's done with him." "You think that's it?" "Look how he's looking out the window, dude." "When he looks out, it's not right." "It's like he's looking past the window." "You guys, there's someone outside." "I was just sitting here, and I saw someone out in the yard, and I thought it was Rob, but then you guys were talking, and the guy was, like, still there." "Oh, for real, Bambers?" "Yeah, for real, Rob." " Should we check it?" "No." "You know, I would so love to not have you always asking me," ""For real, for real?" Yeah, for real!" "Why would I make it up?" "Okay." "I'm not, like, making things up." "You think I'm lying?" "I said it was for real, so it's for real, obviously." "l'm not making it up!" "l'm sorry!" "Guys." "I'm not just saying it just to make shit up!" "God!" "I don't see anything." "Yeah, I don't see anything." "Fuck!" "What the F, man?" "!" "God!" "What did you see, guys?" "Yeah, there was somebody out there, Amber." "We saw a bunch of strange footprints." "There's somebody watching this place, I think." "I get that feeling, too, you know?" "I found a really weird message on the typewriter if you run the ribbon backwards." "I think it's probably just the ding-hole neighbor." "I'm going to bed." "Come on, Amber." "Can I get by?" "God." "Can I?" "God." "is there something you want to tell me, partner?" "Have a good night." "Something hidden?" "ldea -- a website that tells you where something hidden is, things you hate but need, in your area, like new tires or vacuum-cleaner bags." "Those bran breakfast muffins, shit like that." "Mm, poor guy." "Hi, Trevor." "He'll wake up." "I gave him a noctizine." " Yeah, baby, yeah!" " Mm!" "Yeah!" "Yeah, baby!" "Yeah, baby, yeah." "Righteous." "Trevor G rocks my world." "Yeah!" "Yeah, baby!" "Yeah." "Righteous." "Ow." "Oh, my goodness..." "Jesus and Joseph." "Mmm, nummy." " What's up?" " You guys got to look at this." "Okay." "Okay." "Crap." "He hurt her." "Should we tell someone about this?" "Just watch, guys." "That's what the Internet's for." "Rob, I swear to God, you are so twisted." "What?" "Why's everybody against me here?" "I can't believe I'm still hungry." "Here, have some Pop-Tarts." "We're busy watching this." "I don't want Pop-Tarts." "I want a frickin' McMuffin." "BFD, spaz!" "So go get one!" "Christ!" "Guys!" "I should write a song about how big computers are in our lives." "Oh, my God!" "That's OOC!" "Who loves computers more than you?" "My brother!" "Oh, man.. ." "What happened?" "What happened?" "Dah." "There's someone at the door." "I think it's the police checking for a gas leak." "You mean the fuzz?" "Yeah, he must have caught his chemicals on some kind of sensor." "Come on, man." "Find her." "Hurry." " Guys." " Oh.. .my.. .crap." "This is insane." "Dudes, what if this is real?" "Then it's awesome." "Rob, you're like a person I don't even know sometimes." "Oh, yeah?" "Who cares?" "Guys." "It's got to be a joke, though, right?" "Shit's real, Rob." "Some shit's actually real." "Don't get too, like, theoretical." "I'm just saying it doesn't seem that likely that some lady doing regular voyeur, taking her clothes off, whatever, splits and these two terrorists show up and start building bombs and killing people while the cameras are on, does it?" "God." "Think sometimes, Bambo." "Guys, maybe that's it." "Maybe they're trying to terrify us -- l mean, terrorize us -- about how sick they are and how smart and all they are, how many bombs they made." "So, uh, they make up this crap about a naked woman, or they actually get a naked woman -- what better way to attract attention -- and then they move over to this stuff when they got a big audience." "Entertain them terrorize them." "That actually kind of makes a lot of sense." "So, it's about terror." "Big surprise." "What isn't nowadays?" "Bullshit." "September 11th?" "That was real, right, Rob?" "You got to remember, some stuff is, like, actually happening." "Yeah, well, if it's happening, don't worry about it." "Somebody else is probably watching this shit." "Somebody else will figure it out." "Amber, wild wings, come on." "You're not staying here." "I got to crap, Rob." "l think there's something..." "This crap's real." "There's something that we missed." "It's Friday." "We were supposed to be done with this job by now." "Amber!" "Let's roll!" "You need to step off, yo." "Shower, Rob." "Shower, shits, and wedding -- that's my life." "Yeah, so?" "Meat-lover's pizza, "Halo 3 "" "and "Let me cum on your tits" -- that's my life." "Yo." "Rob, I figured something major out." "Oh, yeah?" "What did you figure out, buddy?" "The cylinders, the killer things, are in people -- dead people in coffins being shipped home to be buried." "They dig them up when they get where they're going, and they get out the killer things -- the cylinders." "They put them in tail pipes." "Uh, there are cities marked on every box, and they're going everywhere, 'cause they're impossible to trace through the coffins and through the people in the boxes." "Oh, man..." "Cars are the killers." "And everybody dies." "Okay." "Oh, man." "What?" "Well, everybody's gonna die." "Trevor!" "What's the most pervasive thing you think about?" "What do you mean?" "Like work?" "No." "Like for me, it's being afraid." "You know what I mean?" "No." "Afraid life's just gonna be more of the same." "Like..." "Mother!" "Hey." "So, what did we miss on the site?" "What's up, douche?" "That guy's called Ziploc." "Hmm." "Are you mad at me?" "I'm mad at myself." "I act like such a kid." " What'd I miss?" "Cleaning the place out." "That's Ziploc." "Nice name." "Maybe we'll get to see if that guy finally gets offed." "This is the best shit ever on the Internet." "Oh." "He feels bad." "Poor guy." "'Cause they're gonna do him." "Rob, 'cause he made something that's gonna kill people, probably." "Or not." "I mean, why is he crying?" "That's weird behavior, if they're filming this to, like, try and terrorize people." "Mm..." "He looks depressed." "I know how bad he feels." "Look, he's doing something up there, see?" "Hmm." "What I love about Internet shows is no ads." "I hope that's not the end of the story, 'cause that shit was weak as crap." "Although half the time, I love ads." "I don't know." "I'm going for a run." "I need to work off some of this Monterey Jack, for Christ sakes." "Oh, I guess I'll go with." "William Zill." ""Basement ceiling." "Red key."" "We're in "Four Boxes."" "What's up?" "Rob, you're cracked." "Rob, where are you guys?" "Keeping it real, Amber!" "You guys got to get back here right fucking now." "Uh-huh." ""Four Boxes " uh.. ." "It was a fake." "It was a set." "They took it down." "You won't believe this." "You're not gonna believe this, dude." "It's a house." "It's this house." "We're inside of "Four Boxes."" "Right." "And I'm really freaked out." "You guys got to get back here right fucking now." "I guess they fucking taped it earlier." "It's this place, hombre!" "N.J. -- no joke." "I found the cash in the suitcase." "I -- which I guess is what Ziploc was looking for at the end." "Yeah, Rob." "Fucking real money hidden in the ceiling." "And if the cash is real, then Ziploc's real, and he's gonna be coming 'cause of the fucking note in the typewriter ribbon." "So, you guys have to get back here right fucking now." "Okay, okay, okay." "Um, if the -- lf the MF'ing website was filmed here, then, um, the cameras are there and everywhere." "So, the action we were watching wasn't live." "It was taped." "And Havoc didn't know he was being filmed." "Or he did." "Why would Havoc film this?" "To terrorize people." "He's a fucking terrorist." "He's a fucking terrorist." "We're in Havoc's house." "Oh, my God." "Ziploc in this house." "Oh, fuck me!" "Rob!" "I'm scared." "Okay, okay." "Good." "Um..." ""Silent/Microphone." Rob switched that when we got here." "What's that?" ""Live." "Webcast."" "We've been watching a webcast?" "Okay, switch it to "live."" "Oh, fuck me!" "Send the police right now." "It's an emergency." "I'm at, um, um, William Zill's house, um, at Deer Run Lane, and there's, uh, a Middle Eastern man wearing a turban, and, um, he's got no face," "and he's -- he's, um... uh, he's a terrorist!" "So, come now." "Please!" "God!" "Hurry!" "He's a fucking terrorist!" "Oh, fuck." "Don't come down here." "Come on, come on, come on." "Fuck this!" "Okay." "Here we go." "Trevor?" "Trevor?" "Trevor!" "You got to answer us right now, man!" "What's that smell?" "Amber, let's burn." "Something's wrong." "No!" "Rob, we gotta find Trevor." "No, we don't gotta..." "You still love him, don't you?" "We're not the kind of people who live happily ever after, are we?" "No." "Rob!" "What was that?" "Oh, my God!" "Oh, my God!" "This...isn't...good." "Hey, guys!" "That's a MF'ing wrap!" "Yeah!" "Yeah." "Yeah, baby!" "Yeah, baby." "That's what I'm talking about." "We are gonna be huge stars, thanks to you, Trevor." "Let's get our little butts down to that Friday's and start celebrating." "We did it, man." "I am fucking bursting with wedding info for you, Trevor." "Good morning." "Yep." "Just uploading footage on the URL fourboxes.tv." "Time till completion -- 2 minutes." "You know, this really is the perfect TV show for the Internet." "Just one, long, stupid story in one room." "Rob, quit saying that, for real." "It's my script." "Don't say that anymore." "Well, it's pretty damn stupid, especially my part." "No, it's not stupid." "It's pretty perfect." "Dude, I didn't mean it to sound so harsh." "I'm just saying -- what won't people watch?" "Okay. "Begin webcast at URL fourboxes.tv?"" "Yes, please." "Hey, here's what I wrote earlier " ""Hey, guys." "Check out this website." "It's called fourboxes.tv." "This guy, we call him Havoc, he moved into an apartment set up for voyeur, but he don't know it." "What's cooler still, he's weird, always doing science on mannequins, and talking to some terrorist dude named Ziploc." "We think they're gonna kill someone." "Enjoy."" "Check this shit out." "40 hits." "80 hits..." "Let's get this dude's stuff moving for real now, what we didn't wreck for props." "Six days till we're on live." "Oh, shit!" "Trevor, check it out." "This guy's seen "four Boxes."" "Hmm." "What's "four Boxes"?" "I would do a threesome, I think." "No dudes, though." "10 years since Diana died." "Wow." "That went fast." "Hey, Trevor!" "What?" "238,000 hits." "Can you believe that shit?" "l love it, man." "I love it!" "Maybe it wasn't copacetic for us to come up here, or down here, or wherever we are, you know what I mean, and, like, have everything be all..." "Poor Bill Zill." "God, that's so sad!" "Seems like more than a week ago that we were putting these cameras up." "This place was an awesome backdrop." "Can't believe it took us four months to find, though." "Fuck!" "I got to go to Home Depot." "I can't believe you talked me into this." "You still got a hot bod, babe." "What's the problem?" "I'm flat." "I know guys don't care, but..." "But I know this project was something you wanted to get -- we wanted to get out of the way before our next chapter." "Before we give up for good?" "Sell privacy fencing for the rest of our lives?" "Lots of people get married and have jobs and do the kid thing and still do cool stuff, Trevor." "That's a lie, Amber." "You're so far inside the box, it's just hilarious." "They're all upstairs." "They're all hiding in the bathroom, you fucking idiots!" "rhere was something strange going on in this house." "There's something strange going on inside every house, yo." "Half a million hits, dude." "Half a million hits, dude." "Mm-hmm!" "Rehearsal!" "All right." "Rehearse, rehearse, rehearse." "We got to rehearse the ending." "Tonight's the night we switch to live, and there's gonna be a gang of people watching." "I'm into it." "I guess I am pretty nervous about tonight, actually." "I think I might take a Paxil." "No." "Hey, I will if I want to." "Guys." "Tonight's the night Ziploc kills us, and then we go live." "Trevor G!" "G money!" "Why don't you go first?" "All right." "Uh, so, end of feed, and then, um... .. .surprise!" "No terrorists!" "No murderers!" "Just us, yo!" "Uh, what's up, everybody?" "Y'all boys and girls, how's it going?" "And, um..." "What you all just been watching is a great big show put on by myself, Trevor Grainger, along with my pals Amber Croft, Rob Rankrus, which are, together, what comprises FOURBOXES.TV, L.L.C." "Uh, I'm Trevor Grainger, as I mentioned, and I played Havoc and Ziploc and a city official and other parts in this drama that you've just been watching." "Oh, and we also sat down and decided to call it a cyberdrama." "So, uh, Amber, take it away." "I'm Amber, Amber Croft also known as Thistle." "Um, I played myself," "Havoc's victim, uh, man in the yard, movers, and other parts, and I'm a singer/songwriter and actress." "Oh, and you guys should click to my website," "You can't plug your own website, Amber!" "Shut up!" "You can't talk to me like that!" "I was just gonna say -- if you would have let me finish, I was gonna say that you can link to all of our websites through fourboxes.tv, if you would have let me." "God." "Anyway.. ." "Also, of course, there's lots of cool goodies for sale, too." "Wouldn't be a show, or cyberdrama, without goodies." "Such cool stuff." "Um, you can order a DVD of "Four Boxes "" "uh, cool t-shirt, you can see more racy pictures of me, oh, buy my CD, "Broken Down in the Middle of Nowhere "" "um, and just find out more about me and Trevor and Rob." "And then I'll be like," ""Peace!" "Over to you, Rob " or something." "Right." "Right." "And I'm like, "Hey, everybody." "What's up?" "I'm Rob Rankrus." "I played myself, of course," "Havoc, Ziploc, movers, and other good stuff." "I hope you guys liked our story." "I'm sort of the technical dude behind the operation." "I made it possible for you guys to watch, and I am so happy that you did." "You guys kick ass!" "Yeah." "Right on." "Yeah, you guys rule!" "Yeah, like, buy some stuff from our site." "You know the address, fourboxes.tv." "We're doing a live chat after the cyberdrama, so stick around for that." "Also will be some other good stuff." "That was pretty all right." "I think the main priority will be, you know, to see who puts us on the TV." "So, I think we should stay up live for, like, four hours, then get back, get ready to do the merch." "Once the story goes to TV, that's when the site pops the hardest." "I don't know if I'm gonna go through with it, you guys." "Trevor, you act like such a fucking dick!" "Good." "Work at Allstate forever." "Eat me, fag!" "I'm just trying to get a little reaction from you." "And you get all mental " "Guys..." "Well, what?" "No -- no one's even gonna care!" "All they want to know is did I harrumph both you guys or not." "So, what's so hard about that?" "Tell the truth -- you did me, you grinded him, right?" "Only I really know that, right?" "I'm gonna be like, I'm gonna be like," "# Goodbye, Allstate #" "Right?" "I'm like, "Fuck you, Maple Grove!" "I'm out!"" "I didn't do shit with my hair for tonight." "Hey, Amber, you want to come with me to Buffalo Wild Wings?" "All right." "All right." "Grab one of these boxes." "These are the last of the props in the house." "Proper props, y'all." "Are you coming?" "No. I'm tired." "All right, homes." "Well, is everything cleaned up, then?" "No trace of nothing, man -- just the cameras and us." "Sweet." "Meet you in the CR-V." "I've had something really important to tell you for the last couple days, but I just never feel like you would care." "And you don't." "# Ooh #" "# Ooh #" "# Ooh #" "# Ooh #" "We're in "four Boxes."" "Rob, where are you guys?" "You guys gotta get back here right now." "We're in "four Boxes."" "Ugh, I don't feel so good." "You're never gonna believe this, dude." "It's a house." "Ugh, I feel kind of... lt's this house." "We're inside of "four Boxes."" "I guess I should have eaten something with that Paxil." "Vicodin." "Me, too." "What about our show?" "Hey." "We're in Havoc's house." "Rob!" "I'm scared." "Okay, okay." "Good." "I want to be a Red..." "Hot..." "Chili Pepper." "911 , what's your emergency?" "Send the police right now." "It's an emergency." "Um, William Zill's house, um, at Deer Run Lane." "Are you on Deer Run Lane or on Deer Run Circle?" "Hurry!" "He's a fucking terrorist!" "Calm down, sir." "We have officers coming right now." "Trevor?" "You Qot to answer us right now, man." "Trevor, baby?" "Something's wrong." "No, Rob." "We got to find Trevor." "No, we don't gotta.. ." "Trevor, baby I have something important to tell you." "You still love him, don't you?" "We got the pagoda." "We're not the kind of people who live happily ever after, are we?" "Mo." "Rob, what was that?" "!" "Oh, my God!" "This...isn't...good." "# All we ever talk about #" "# Is all the things we think about #" "# Like love and pain and loneliness #" "# And me and you and tenderness #" "# And now we come to the place we start #" "# Living to be free #" "# Looking out for me #" "# Rise above this eventually #" "# Here I am #" "# On the dark side of the planet #" "# In the far side of space #" "# There's a universe of stars #" "# And someone to take your place #" "# And love can fill #" "# The emptiness between #" "# So the drift will come together #" "# And we'll live to be free #" "# Looking out for you and me #" "# Gonna rise above this eventually #" "# And now I'm on my own again #" "# Learning how to love again #" "# My fate will find me eventually #" "# My heart struggles on #" "# So spiritually #" "# And I'll get to the place #" "# Where I'm gonna start #" "# Living to be free #" "# Looking out for me #" "# Gonna rise above this eventually #" | Low | [
0.47038327526132406,
33.75,
38
] |
The Boy Who Came In From The Cold Title: The Boy Who Came In From The ColdAuthor: B.G. ThomasCover Artist: Aaron AndersonPublisher: DreamspinnerBuy Link:Amazon.com The Boy Who Came in from the ColdGenre: M/M ContemporaryLength: NovelRating:5 stars out of 5 A Guest Review by Sammy Review Summary: A novel that travels beyond the idea of a “good samaritan” and revels in the idea that to “pay it forward” is much more than a quaint cliche. Blurb: Todd Burton has had enough of small-town Buckman. His abusive stepfather calls him a fag; his friend Austin makes him realize he may be gay, but Todd doesn’t want to admit his stepfather is right; and he dreams of being a chef. Three good reasons to flee his hometown and pursue greener pastures. But when Todd reaches the big city, his luck runs dry. Soon he can’t pay his rent and gets evicted. In the middle of a snowstorm. Gabe Richards is a wealthy businessman with enough wounds of his own to make him afraid of ever being intimate again. But when he sees Todd outside his building, freezing to death, he takes pity on him and takes him in from the cold. To their mutual surprise, Todd and Gabe find themselves drawn to each other. “One night” turns into a week. Maybe letting a man in from the cold can melt the ice around Gabe’s heart—and maybe getting evicted will turn Todd’s luck around. Review: “But please remember this also, if nothing else. Remember the words of Albert Einstein: ‘There are two ways to live your life. One is as though nothing is a miracle. The other is as though everything is a miracle.'” I will admit that I did not set out thinking this novel was going to be a five star read. I was worried, you see, worried that it might be a story based on some old tired cliche. You know, we’ve all read them: abused small town boy comes to the city with big dreams only to have them dashed…homeless, he finds a truer than life good samaritan and they fall hopelessly in love. So, you see, I was totally unprepared for this novel, The Boy Who Came In From The Cold by B.G. Thomas. It was far from tired, and certainly no fulfilled cliche. Instead, it was poetry. A lyrical dance that carried me far beyond the story into the heart and mind of its heroes. Yes, this novel was, at its heart, a lesson in how when a person decides to give back–to “pay forward” some of the charity that had been shown to them in their past, they will actually reap more. And, yes, at its core, it held the hint of a good samaritan parable; one man reaching out to help a stranger, to lighten his load and care for him in his need. But, you see, this story was so much more than that. B. G. Thomas takes us on a journey. He goes inside the mind of a young man who has denied who he is for most of his life. Fighting back years of abusive behavior from his stepfather and the heartbreak of a weak and uncaring mother, Todd Burton has more, much more to fight than the cold and his poverty. He has to fight the man within who cannot allow himself to be what he is…a man who is attracted to other men…a boy who has repeatedly fought off the the idea that he could be gay. However, that is not what made this story uniquely different. Please, I do not mean for this to sound unkind or cast aspersions on other stories that may have a similar themes, but what sets this author’s novel apart is how very intelligently it is written. B.G. Thomas takes what is an admittedly tried and oft told story line and infuses it with provocative language and characters that push against the stereotype. Todd and Gabe don’t “fall in love”. Instead, they fight against it with almost every breath. One because he must deny who he truly is and the other because he has made the mistake of “saving” someone before and painfully lost his heart in the process. The author keeps us on the edge of our seat waiting for Todd to run away again, for Gabe to place yet another brick in the wall he has built around himself. Along the way, he tosses in quotes from notable sources juxtaposed against his own clever prose to remind us of our own frail humanity. The Boy Who Came In From The Cold does not just encourage us to keep an eye out for the one who is in need, it reminds us that we all will at one time or another be that man in need. This story makes the bold move to believe there is hope in a jaded world of where men and women fight for basic rights; that there is within us the ability to be a “good samaritan”, if we only take the risk. I highly recommend The Boy Who Came In From The Cold by B.G. Thomas. It is a five star read! Like this: Related Author Sammy A mature woman, gracefully growing older, who lives with 12 cats and talks to imaginary people–had ya going there for a minute didn’t I? I am an avid lover of all things m/m who delights in occasionally teasing Wave! 10 comments Excellent review of an excellent book. Kept rereading it to see what is so different about this story because it could be so been here, read this. It is the prose and the pacing. Sit down to read it and it’s done. Such flow sucks you in and does not deviate or stutter. Loved this book! Oh, to be able to write like this! I liked this book, too, but I saw it a little differently. I thought it was a case of “insta-love”, and there was quite a fairy-tailish quality to the story. I didn’t mind that though. (I also enjoyed Todd’s love of and talent for cooking. The recipes he came up with sounded delicious.) Gaycrow–I definitely saw those elements as well–but the chord it struck with me resonated a bit differently. I love that we all can find varied elements that appeal to us in this author’s writing–a sign of a good story! 😀 Sammy, this proves why reviews are so important, because I got a similar first impression with this book, and I didn’t buy it. Now, after reading your review, I’m going to get the book, thanks. BTW Excellent review! | High | [
0.669410150891632,
30.5,
15.0625
] |
Q: definition of the tensor product of dual spaces I have $V$ a real finite dimensional vectorspace. I would like to know what exactly $V^*\otimes \cdots \otimes V^*$ ($r$ times) is. I know that given a vectorspace $W$, $W\otimes W$ is just the vectorspacer spanned by the elements $e_i\otimes e_j$ where $\{e_i\}$ is a basis of $V$. However, $V^*$ is not just vectorspace, it also has an action on $V$ so I would expect $V^*\otimes \cdots \otimes V^*$ to also have an action on $V^r$ (or on $V\otimes \cdots \otimes V$ maybe). How is this defined? I would say $$(\alpha_1\otimes\cdots\otimes\alpha_r)(v_1,\cdots,v_r) = \sum\alpha_i(v_i)$$ But I don't know if that's it A: The tensor product of (multi)linear forms is defined by the product of their images, i.e., $$ (\alpha_1\otimes\cdots\otimes\alpha_r)(v_1,\cdots,v_r) = \prod_{i=1}^r\alpha_i(v_i) $$ You can look up more definitions of tensor products here. | High | [
0.674157303370786,
37.5,
18.125
] |
On This Page Fallback / Retry Strategy This topic explains the need for a fallback/retry strategy for Dynamic Ingest requests, and provides a high-level description of how to implement one. Background Dynamic Ingest imposes rate-limiting (per account) in two ways: No more than 20 requests (CMS API and/or Ingest API requests) per second are allowed No more than 100 concurrent normal priority jobs are allowed The first is not difficult to manage in your app - you can just impose a delay of 3 seconds or more between requests. The second is more complicated, as there is no way to query the system directly to determine how many jobs you have currently processing. Alternatively, consider using the Low Priority queue which will allow you to queue more than 100 jobs. You can simply wait for some period of time and retry requests until they succeed, but you can implement a more rational fallback/retry system by listening for notifications from the Dynamic Ingest system and using the information to keep track of jobs in-flight yourself. One way to implement this would be to build a transceiver app that both submits the ingest requests and listen for notifications. The diagram below shows the high-level logic of such an app. | Mid | [
0.6355140186915881,
42.5,
24.375
] |
Paril Saddle Paril Saddle (Parilska Sedlovina \pa-'ril-ska se-dlo-vi-'na\) is a saddle of elevation 1398 m in Friesland Ridge in Tangra Mountains on Livingston Island, Antarctica. Bounded by St. Boris Peak to the northeast, and Simeon Peak to the southwest. Overlooking Macy Glacier to the east and southeast, and Huntress Glacier to the west and northwest. Bulgarian topographic survey in 1995/96 (estimated elevation 1390 m), and mapping in 2005 and 2009. It was first visited and GPS-surveyed by the Bulgarian climbers D. Boyanov, N. Petkov and N. Hazarbasanov on 15 January 2017. The feature is named after the settlement of Paril and the adjacent homonymous saddle between Pirin Mountain and Slavyanka Mountain in southwestern Bulgaria. Maps L.L. Ivanov et al. Antarctica: Livingston Island and Greenwich Island, South Shetland Islands. Scale 1:100000 topographic map. Sofia: Antarctic Place-names Commission of Bulgaria, 2005. L.L. Ivanov. Antarctica: Livingston Island and Greenwich, Robert, Snow and Smith Islands. Scale 1:120000 topographic map. Troyan: Manfred Wörner Foundation, 2009. Antarctic Digital Database (ADD). Scale 1:250000 topographic map of Antarctica. Scientific Committee on Antarctic Research (SCAR). Since 1993, regularly upgraded and updated. L.L. Ivanov. Antarctica: Livingston Island and Smith Island. Scale 1:100000 topographic map. Manfred Wörner Foundation, 2017. Notes References Paril Saddle. SCAR Composite Gazetteer of Antarctica. Bulgarian Antarctic Gazetteer. Antarctic Place-names Commission. (details in Bulgarian, basic data in English) External links Paril Saddle. Copernix satellite image Category:Mountain passes of Livingston Island Category:Bulgaria and the Antarctic | High | [
0.677640603566529,
30.875,
14.6875
] |
Q: update date status bar icon in android I develop an app that shows date in status bar as icon(an image for every day,1,2,3,...),now my problem is that when the date change i can't sense it. is there any way that can catch the change of date and update the status bar icon. mNM =(NotificationManager) getSystemService(NOTIFICATION_SERVICE); int resId = getResources().getIdentifier("p_"+Day_Num, "drawable", getPackageName()); Notification notification = new Notification(resId, "", System.currentTimeMillis()); RemoteViews contentView = new RemoteViews(getPackageName(), R.layout.custom_notification); contentView.setImageViewResource(R.id.image, resId); contentView.setTextViewText(R.id.title, Day); contentView.setTextViewText(R.id.text, Date); contentView.setOnClickPendingIntent(R.id.imageView1, PendingIntent.getActivity(this, 0, new Intent(this, Option.class),0)); notification.contentView = contentView; Intent notificationIntent = new Intent(this, Option.class); PendingIntent contentIntent = PendingIntent.getActivity(this, 0, notificationIntent, 0); notification.contentIntent = contentIntent; notification.defaults = Notification.FLAG_NO_CLEAR; mNM.notify(NOTIFICATION, notification); when I use servic Calservice.java: public void onCreate() { //code to execute when the service is first created alarmManager = (AlarmManager) getSystemService(ALARM_SERVICE); Intent intent = new Intent(ALARM_REFRESH_ACTION); pendingIntent = PendingIntent.getBroadcast(this, 0, intent, PendingIntent.FLAG_CANCEL_CURRENT); alarmReceiver = new BroadcastReceiver() { @Override public void onReceive(Context context, Intent intent) { alarmCounted++; Message msg = myHandler.obtainMessage(ALARM_CODE, intent); myHandler.sendMessage(msg); } }; IntentFilter filter = new IntentFilter(ALARM_REFRESH_ACTION); registerReceiver(alarmReceiver, filter); mNM =(NotificationManager) getSystemService(NOTIFICATION_SERVICE); utils.getCurrentShamsidate(); startRepeating(); } public void startRepeating() { // We get value for repeating alarm int startTime =1000; long intervals =1000; // We have to register to AlarmManager Calendar calendar = Calendar.getInstance(); calendar.setTimeInMillis(System.currentTimeMillis()); calendar.add(Calendar.MILLISECOND, startTime); // We set a repeating alarm alarmManager.setRepeating(AlarmManager.RTC, calendar .getTimeInMillis(), intervals, pendingIntent); } private void showNotification() { mNM.cancelAll(); utils.getCurrentShamsidate(); int resId = getResources().getIdentifier("p_"+Day_Num, "drawable", getPackageName()); Notification notification = new Notification(resId, "", System.currentTimeMillis()); RemoteViews contentView = new RemoteViews(getPackageName(), R.layout.custom_notification); contentView.setImageViewResource(R.id.image, resId); contentView.setTextViewText(R.id.title, Day); contentView.setTextViewText(R.id.text, Date); contentView.setOnClickPendingIntent(R.id.imageView1, PendingIntent.getActivity(this, 0, new Intent(this, Option.class),0)); notification.contentView = contentView; Intent notificationIntent = new Intent(this, Option.class); PendingIntent contentIntent = PendingIntent.getActivity(this, 0, notificationIntent, 0); notification.contentIntent = contentIntent; mNM.notify(NOTIFICATION, notification); } public void doscan() { String Day = Day_Num; utils.getCurrentShamsidate(); if(!Day_Num.equals(Day)) showNotification(); } A: @hamhame, Why don't you try an AlarmManager, to set alarm on Every morning at 12:00 Am. in this way you can set your opration and do things pretty easily.. This link can help you more on how to set a repeating alarm for Time interval of 24 hours. for listening to TIme/Date change by user see this answer. you need to register a receiver(with required action given in above link) via manifest (if want to listen always ) or register in Activity/service according to your need. and in receiver you will got to know what time has been set by getting current time and can modify you AlarmReceiver too. | Mid | [
0.5717761557177611,
29.375,
22
] |
--- abstract: 'Generative adversarial nets (GANs) have become a preferred tool for accommodating complicated distributions, and to stabilise the training and reduce the mode collapse of GANs, one of their main variants employs the integral probability metric (IPM) as the loss function. Although theoretically supported, extensive IPM-GANs are basically comparing moments in an embedded domain of the *critic*. We generalise this by comparing the distributions rather than the moments via a powerful tool, i.e., the characteristic function (CF), which uniquely and universally contains all the information about a distribution. For rigour, we first establish the physical meaning of the phase and amplitude in CFs. This provides a feasible way of manipulating the generation. We then develop an efficient sampling way to calculate the CFs. Within this framework, we further prove an equivalence between the embedded and data domains when a reciprocal exists, which allows us to develop the GAN in an auto-encoder way, by using only two modules to achieve bi-directionally generating clear images. We refer to this efficient structure as the reciprocal CF GAN (RCF-GAN). Experimental results show the superior performances of the proposed RCF-GAN in terms of both generation and reconstruction.' author: - | Shengxi Li\ Department of Electrical and Electronic Engineering\ Imperial College London\ `[email protected]`\ Zeyang Yu\ Department of Electrical and Electronic Engineering\ Imperial College London\ `[email protected]`\ Min Xiang\ Department of Electrical and Electronic Engineering\ Imperial College London\ `[email protected]`\ Danilo Mandic\ Department of Electrical and Electronic Engineering\ Imperial College London\ `[email protected]`\ title: Reciprocal Adversarial Learning via Characteristic Functions --- Introduction ============ Generative adversarial nets (GANs) have achieved impressive success by virtue of its powerful capability in capturing complicated data distributions [@goodfellow2014generative]. In practical applications, however, their significant potential still remains under-explored as GANs typically suffer from unstable training and mode collapse issues [@mescheder2018training]. An effective yet elegant way to address these issues is to replace the Jensen-Shannon (JS) divergence in measuring the discrepancy in the original form of GANs [@Arjovsky2017TowardsPM] by another class of metrics called the integral probability metric (IPM) [@muller1997integral] given by, $$\label{ipm-gan-general} d(\mathcal{P}_d, \mathcal{P}_g) = \sup_{f\in\mathcal{F}}|\mathbb{E}_{x\sim\mathcal{P}_d}[f(x)] - \mathbb{E}_{x\sim\mathcal{P}_g}[f(x)]|,$$ where the symbol $\mathcal{F}$ in IPMs represents a collection of (typically real) bounded functions, $\mathcal{P}_g$ denotes the generated distribution, and $\mathcal{P}_d$ is the real data distribution. Using IPMs to improve GANs has been justified by the fact that in real-world data distributions are typically embedded in low-dimensional manifolds, which is intuitive because data preserve semantic information instead of being a collection of rather random pixels. Thus, the divergence measure (“bin-to-bin” comparison) of the original GAN could easily max out, whereas the IPMs such as the Wasserstein distance (“cross-bin” comparison) can consistently yield a meaningful measure between the generated and real data distributions [@Arjovsky2017TowardsPM]. Varying collections of $\mathcal{F}$ in (\[ipm-gan-general\]), therefore, define different IPM-GANs and the supremum $\sup_{f\in\mathcal{F}}$ is then typically achieved by the discriminator net, or more formally, the *critic* in the IPM-GANs. The first IPM-GAN was motivated by the Wasserstein GAN (W-GAN) [@arjovsky2017wasserstein], where $\mathcal{F}$ denotes all the 1-Lipschitz functions. However, it has been widely argued that the *critic* is not powerful enough to search within all the 1-Lipschitz function spaces, which leads to limited diversity of the generator due to an ill-posed equivalence measurement of $\mathcal{P}_d$ and $\mathcal{P}_g$ [@arora2017generalization; @arora2017gans]. Follow-up works have been proposed to improve the W-GAN by either enhancing it to satisfy the 1-Lipschitz condition (e.g., by gradient penalty [@gulrajani2017improved] or spectral normalization [@miyato2018spectral]) or by employing *easy-to-implement* $\mathcal{F}$ for the *critic*. The latter, by virtue of relaxing the *critic*, typically leads to a stringent comparison on the embedded feature domain, i.e., by matching higher-order moments instead of the mean matching in the W-GAN. This path includes many recent GANs of additionally considering the second-order moment (e.g., Fisher-GAN [@mroueh2017fisher] and McGAN [@mroueh2017mcgan]), together with explicitly (e.g., Sphere GAN [@park2019sphere]) or implicitly (e.g., MMD-GAN [@li2017mmd; @binkowski2018demystifying]) comparing higher-order moments. We refer to the *Related Works* section in the supplementary material for a detailed analysis on different IPM-GANs. Generalising as moment matching has been justified as a natural and beneficial way when understanding IPM-GANs [@farnia2018convex; @liu2017approximation; @mohamed2016learning]. This also compensates for the deficiency where the *critic* may not transform the data distributions into unimodal distributions, for example, the Gaussian distribution that is solely determined by the first- and second-order moments. Moreover, it is even more safe to compare the distributions because the equivalence in distributions ensures the equivalence in moments; the inverse, however, does not necessarily hold. As a powerful tool of containing all the information relevant to a distribution, the *characteristic function* (CF) provides a universal way of comparing distributions even when their probability density functions (pdfs) do not exist. The CF also has a one-to-one correspondence to the cumulative density function (cdf), which has also been verified to benefit the design of GANs [@mroueh2017sobolev]. Compared to the moment generating function (mgf) that has been reflected in the MMD-GAN [@li2017mmd], the CF is unique and universally existent. More importantly, the CF is automatically aligned at $\mathbf{0}$; this means that even a simple “bin-to-bin” comparison between CFs can consistently provide a meaningful measure and thus avoid gradient vanishing that appears in the original GAN [@arjovsky2017wasserstein]. On the other hand, the weak convergence property of CFs ensures that the convergence in the CF also indicates the convergence in the distributions. ![The overall structure of the proposed RCF-GAN. The generator serves to minimise the CF loss between the embedded real and fake distributions. The *critic* serves to minimise the CF loss between the embedded real and the input noise distributions, whilst maximising the CF loss between the embedded fake and the input noise distributions. Moreover, an MSE loss between the embedded fake and the input noise distributions is regularised as the auto-encoder loss, which has not been shown in the figure. An optional $t$-net can be employed to optimally sample the CF loss. []{data-label="fig_overall"}](ChF_GAN_Structure.png){width="80.00000%"} In this paper, we propose a reciprocal CF GAN (RCF-GAN) as a natural generalisation of the existing IPM-GANs, with the overall structure shown in Fig. \[fig\_overall\]. It needs to be pointed out that incorporating the CF in a GAN is non-trivial because the CF is basically complex-valued and the comparison has to be performed on functions as well. To address these difficulties, we first demystify the role of CFs by finding that its phase is closely related to the distribution centre, whereas the amplitude dominates the distribution scale. This provides a feasible way of balancing the accuracy and diversity of generation. Then, as for the comparison over functions, we prove that other than in the whole space of CFs, sampling within a small ball around $0$ of CFs is sufficient to compare two distributions. We further optimise the sampling strategy by automatically adjusting sampling distributions under the umbrella of the *scale mixture of normals* [@li2019solving]. Moreover, because we are comparing in the embedded feature domain, we prove its equivalence to the counterpart in the data domain when a reciprocal between the generator and the *critic* holds. This motivates us to incorporate an auto-encoder structure to satisfy this theoretical requirement. Different from many existing adversarial works with auto-encoders incorporating at least three modules[^1] [@li2017mmd; @binkowski2018demystifying; @makhzani2015adversarial; @larsen2015autoencoding; @donahue2016adversarial; @brock2016neural; @che2016mode; @dumoulin2016adversarially], our RCF-GAN only requires two modules that already exist in a GAN; the *critic* is an encoder and the generator is a decoder as well, which is neat and reasonable as this comes without increasing computational complexity and without other requirements such as the Lipschitz continuity. More importantly, by only comparing distributions in the embedded domain, our CF-GAN avoids the smoothing artefact that arises from the use of point-wise mean square error (MSE) employed in the data domain, and thus benefits from both the auto-encoder and the GANs, i.e., *bi-directionally* generating *clear* images. Our experimental results show that by employing a naive structure, our RCF-GAN achieves remarkable improvements on the generation, together with an additional gain in the reconstruction. The performance of our RCF-GAN[^2] can be further improved by many orthogonal techniques, e.g., progressively training [@karras2017progressive], using individual learning rates [@heusel2017gans], spectral normalisation [@miyato2018spectral] or adding noise to discriminators [@Arjovsky2017TowardsPM]. Characteristic Function Loss and Efficient Sampling Strategy ============================================================ Characteristic Function and Elliptical Distribution {#sec_cfintro} --------------------------------------------------- The CF of a random variable $\bm{\mathcal{X}}\in\mathbb{R}^m$ represents the expectation of its complex unitary transform, given by $$\label{eq_chf} \Phi_{\bm{\mathcal{X}}}(\mathbf{t}) = \mathbb{E}_{\bm{\mathcal{X}}}[e^{j\mathbf{t}^T\mathbf{x}}] = \int_\mathbf{x} e^{j\mathbf{t}^T\mathbf{x}}dF_{\bm{\mathcal{X}}}(\mathbf{x}),$$ where $F_{\bm{\mathcal{X}}}(x)$ is the cdf of $\bm{\mathcal{X}}$. We thus have $\Phi_{\bm{\mathcal{X}}}(\mathbf{0}) \!=\! 1$ and $|\Phi_{\bm{\mathcal{X}}}(\mathbf{t})| \!\leq\! 1$ for all $\mathbf{t}$. This property ensures that CFs can be straightforwardly compared in a “bin-to-bin” manner, because all CFs are automatically aligned at $\mathbf{t}\!=\!\mathbf{0}$. Moreover, when the pdf of $\bm{\mathcal{X}}$ exists, the expression in is equal to its inverse Fourier transform; this ensures that $\Phi_{\bm{\mathcal{X}}}(\mathbf{t})$ is uniformly continuous. Another important property of the CF is that it uniquely and universally retains all the information of a random variable. In other words, a random variable does not necessarily need to possess a pdf (e.g., when it is an $\alpha$-stable distribution), but its CF always exists. As the cdf $F_{\bm{\mathcal{X}}}(\mathbf{x})$ is unknown and is to be compared, we employ the empirical characteristic function (ECF) as an asymptotic approximation in the form of $\widehat{\Phi}_{\bm{\mathcal{X}}_n}(\mathbf{t}) = \sum_{i=1}^n e^{j\mathbf{t}^T\mathbf{x}_i}$, where $\{\mathbf{x}_i\}_{i=1}^n$ are $n$ i.i.d samples drawn from $\bm{\mathcal{X}}$. As a result of the *Levy continuity theorem* [@williams1991probability], the ECF converges weakly to the population CF [@feuerverger1977empirical]. More importantly, the *uniqueness theorem* guarantees that two random variables have the same distribution if and only if their CFs are identical [@lukacs1972survey]. Therefore, together with the weak convergence, the ECF provides a feasible and good proxy to the distribution, which has also been preliminarily applied in two sample test [@epps1986omnibus; @chwialkowski2015fast]. Before proceeding further, we introduce an important class of distributions that will be mentioned later. \[examp1\] In unimodal distributions, there is one broad class of distributions called the elliptical distribution, which is general enough to include various typical distributions such as the Gaussian, Laplace, Cauchy, Student-*t*, $\alpha$-stable and logistic distributions. The elliptical distributions do not necessarily have pdfs and we refer to [@fang2018symmetric] for more detail. The CF of an elliptical distribution, $\bm{\mathcal{X}}$, however, always exists and has the following form $$\label{eq_ellipticalchf} \Phi_{\bm{\mathcal{X}}}(\mathbf{t}) = e^{j\mathbf{t}^T{\bm{\mu}}}\psi(\mathbf{t}^T\mathbf{\Sigma}\mathbf{t}),$$ where $\bm{\mu}$ denotes the distribution centre and $\mathbf{\Sigma}$ is the distribution scale, and $\psi(\cdot)$ is a real-valued scalar function, for example, $\psi(s) = e^{(\nicefrac{-s}{2})}$ for the Gaussian distribution. By inspecting we can easily find that the phase of the CF is solely related to the location of data centre and the amplitude is only governed by the distribution scale (diversity). Distance Measure via Characteristic Functions --------------------------------------------- The auto alignment property of the CFs allows us to incorporate a simple “bin-to-bin” comparison over two complex-valued CFs (corresponding to two random variables $\bm{\mathcal{X}}$ and $\bm{\mathcal{Y}}$), in the form $$\label{eq_chfloss} \mathcal{C}_{\bm{\mathcal{T}}}(\bm{\mathcal{X}}, \bm{\mathcal{Y}}) =\int_\mathbf{t} \big(\underbrace{({\Phi}_{\bm{\mathcal{X}}}(\mathbf{t}) - {\Phi}_{\bm{\mathcal{Y}}}(\mathbf{t}))({\Phi}^*_{\bm{\mathcal{X}}}(\mathbf{t}) - {\Phi}^*_{\bm{\mathcal{Y}}}(\mathbf{t}))}_{c(\mathbf{t})}\big)^{\nicefrac{1}{2}}dF_{\bm{\mathcal{T}}}(\mathbf{t}),$$ where ${\Phi}^*$ denotes the complex conjugate of ${\Phi}$ and $F_{\bm{\mathcal{T}}}(\mathbf{t})$ is the cdf of a sampling distribution on $\mathbf{t}$. For the convenience of subsequent analysis, we represent the quadratic term for each $\mathbf{t}$ as $c(\mathbf{t})\! =\! ({\Phi}_{\bm{\mathcal{X}}}(\mathbf{t}) \!-\! {\Phi}_{\bm{\mathcal{Y}}}(\mathbf{t}))({\Phi}^*_{\bm{\mathcal{X}}}(\mathbf{t}) \!-\! {\Phi}^*_{\bm{\mathcal{Y}}}(\mathbf{t}))$. More importantly, $\mathcal{C}_{\bm{\mathcal{T}}}(\bm{\mathcal{X}}, \bm{\mathcal{Y}})$ is a valid distance that measures the difference of two random variables via CFs, of which the proof is provided in Lemma \[lemma\_chfloss\_distance\]; this means $\mathcal{C}_{\bm{\mathcal{T}}}(\bm{\mathcal{X}}, \bm{\mathcal{Y}})\!=\!0$ if and only if $\bm{\mathcal{X}}\!=^d\!\bm{\mathcal{Y}}$. A specific type of $\mathcal{C}_{\bm{\mathcal{T}}}(\bm{\mathcal{X}}, \bm{\mathcal{Y}})$ in is when the pdf of $\mathbf{t}$ is proportional to ${||\mathbf{t}||}^{-1}$, and its relationship to other metrics, including the Wasserstein and Kolmogorov distances, has been analysed in detail [@bobkov2016proximity]. Please note that in this paper, we move all the proofs to the supplementary material. \[lemma\_chfloss\_distance\] The discrepancy between $\bm{\mathcal{X}}$ and $\bm{\mathcal{Y}}$, given by $\mathcal{C}_{\bm{\mathcal{T}}}(\bm{\mathcal{X}}, \bm{\mathcal{Y}})$ in , is a distance metric when the support of $\bm{\mathcal{T}}$ resides in $\mathbb{R}^m$. Furthermore, as the phase and amplitude of a CF indicate the data centre and diversity, we inspect $c(\mathbf{t})$ and rewrite it in a physically meaningful way, i.e., through the differences in the corresponding phase and amplitude terms as [@yu2019widely], $$\label{eq_chfloss_deomposed} \begin{aligned} c(\mathbf{t}) &= |{\Phi}_{\bm{\mathcal{X}}}(\mathbf{t})|^2 + |{\Phi}_{\bm{\mathcal{Y}}}(\mathbf{t})|^2 - {\Phi}_{\bm{\mathcal{X}}}(\mathbf{t}){\Phi}^*_{\bm{\mathcal{Y}}}(\mathbf{t})-{\Phi}_{\bm{\mathcal{Y}}}(\mathbf{t}){\Phi}^*_{\bm{\mathcal{X}}}(\mathbf{t})\\ &= |{\Phi}_{\bm{\mathcal{X}}}(\mathbf{t})|^2 + |{\Phi}_{\bm{\mathcal{Y}}}(\mathbf{t})|^2 -|{\Phi}_{\bm{\mathcal{X}}}(\mathbf{t})||{\Phi}_{\bm{\mathcal{Y}}}(\mathbf{t})|(2\cos(\mathbf{a}_{\bm{\mathcal{X}}}(\mathbf{t}) - \mathbf{a}_{\bm{\mathcal{Y}}}(\mathbf{t}))) \\ &= |{\Phi}_{\bm{\mathcal{X}}}(\mathbf{t})|^2 + |{\Phi}_{\bm{\mathcal{Y}}}(\mathbf{t})|^2 - 2|{\Phi}_{\bm{\mathcal{X}}}(\mathbf{t})||{\Phi}_{\bm{\mathcal{Y}}}(\mathbf{t})|+ 2|{\Phi}_{\bm{\mathcal{X}}}(\mathbf{t})||{\Phi}_{\bm{\mathcal{Y}}}(\mathbf{t})|\big(1 -\cos(\mathbf{a}_{\bm{\mathcal{X}}}(\mathbf{t}) - \mathbf{a}_{\bm{\mathcal{Y}}}(\mathbf{t}))\big)\\ &= \underbrace{(|{\Phi}_{\bm{\mathcal{X}}}(\mathbf{t})| - |{\Phi}_{\bm{\mathcal{Y}}}(\mathbf{t})|)^2}_{\textrm{amplitude~difference}} + 2|{\Phi}_{\bm{\mathcal{X}}}(\mathbf{t})||{\Phi}_{\bm{\mathcal{Y}}}(\mathbf{t})|\underbrace{(1 - \cos(\mathbf{a}_{\bm{\mathcal{X}}}(\mathbf{t}) - \mathbf{a}_{\bm{\mathcal{Y}}}(\mathbf{t})))}_{\textrm{phase~difference}}, \end{aligned}$$ where $\mathbf{a}_{\bm{\mathcal{X}}}(\mathbf{t})$ and $\mathbf{a}_{\bm{\mathcal{Y}}}(\mathbf{t})$ represent the angles of ${\Phi}_{\bm{\mathcal{X}}}(\mathbf{t})$ and ${\Phi}_{\bm{\mathcal{Y}}}(\mathbf{t})$, respectively. Therefore, we can clearly see that $\mathcal{C}_{\bm{\mathcal{T}}}(\bm{\mathcal{X}}, \bm{\mathcal{Y}})$ basically measures the amplitude difference and the phase difference weighted by the amplitudes. We can further add a convex combination of the two terms via $0\!\leq\!\alpha\!\leq1$: $$\label{eq_chfloss_deomposed_2} c_\alpha(\mathbf{t}) = \alpha\big((|{\Phi}_{\bm{\mathcal{X}}}(\mathbf{t})| - |{\Phi}_{\bm{\mathcal{Y}}}(\mathbf{t})|)^2\big) + (1-\alpha)\big(2|{\Phi}_{\bm{\mathcal{X}}}(\mathbf{t})||{\Phi}_{\bm{\mathcal{Y}}}(\mathbf{t})|(1 - \cos(\mathbf{a}_{\bm{\mathcal{X}}}(\mathbf{t}) - \mathbf{a}_{\bm{\mathcal{Y}}}(\mathbf{t})))\big).$$ Recall that for the elliptical distributions in Example \[examp1\], the phase represents the distribution centre while the amplitude represents the scale; $\mathcal{C}_{\bm{\mathcal{T}}}(\bm{\mathcal{X}}, \bm{\mathcal{Y}})$ thus measures the both discrepancy of the centres and diversity of two distributions. We show in Figure \[figure\_twoexp\]-(a) that by swapping the phase and amplitude parts, the saliency information follows the phase part of the CF, which captures the centres of the distribution. We further illustrate in Figure \[figure\_twoexp\]-(b) that this property still holds in real data distributions, even though they are much complicated and even non-unimodal. From Figure \[figure\_twoexp\]-(b), solely training the phase results to generating images similar to an average of the real data, as a result of minimising the difference on the data centres. On the other hand, when training the amplitude only, we can obtain diversified but inaccurate images (some digits are wrongly generated). Therefore, by using different weights in $c_\alpha(\mathbf{t})$, we can flexibly capture the main content via minimising the phase difference, whilst enriching the diversity of generated images by increasing the amplitude loss. This provides a meaningful and feasible way of understanding the GAN loss in controlling the generation. The term $\int_{\mathbf{t}}c(\mathbf{t})dF_{\bm{\mathcal{T}}}(\mathbf{t})$ contained in our CF loss can also be employed to optimise the form of kernels via optimising $F_{\bm{\mathcal{T}}}(\mathbf{t})$, whereas the kernels are fixed in the MMD-GANs. This is due to the fact that the shift invariant and characteristic kernels in the MMD metric have to satisfy $k(\mathbf{x},\mathbf{y}) = \int_{\mathbf{t}}e^{-j\mathbf{t}^T(\mathbf{x}-\mathbf{y})}dF_{\bm{\mathcal{T}}}(\mathbf{t})$ for some compactly supported $F_{\bm{\mathcal{T}}}(\mathbf{t})$ [@sriperumbudur2010hilbert]. Thus, the elliptical distributions in Example \[examp1\] potentially provide a set of well-defined characteristic kernels, by choosing $F_{\bm{\mathcal{T}}}(\mathbf{t})$ as a normalised version of the CFs in . Then, the corresponding real-valued kernels are the density generators in [@li2019solving]. We introduce in the next section a way of optimising $F_{\bm{\mathcal{T}}}(\mathbf{t})$. Sampling the Characteristic Function Loss ----------------------------------------- In practice, to calculate $\mathcal{C}_{\bm{\mathcal{T}}}(\bm{\mathcal{X}}, \bm{\mathcal{Y}})$ efficiently, as mentioned in Section \[sec\_cfintro\], ${\Phi}_{\bm{\mathcal{X}}}(\mathbf{t})$ and ${\Phi}_{\bm{\mathcal{Y}}}(\mathbf{t})$ can be evaluated by the ECFs of $\bm{\mathcal{X}}$ and $\bm{\mathcal{Y}}$, which are weakly convergent to the corresponding population CFs. The remaining task is to sample from $F_{\bm{\mathcal{T}}}(\mathbf{t})$. A direct approach is to use the neural net where the input is Gaussian noise and the output is the samples of $F_{\bm{\mathcal{T}}}(\mathbf{t})$. However, Proposition \[prop\_sampling\] indicates that this can lead to ill-posed optima whereby $F_{\bm{\mathcal{T}}}(\mathbf{t})$ converges to some point mass distributions. In our experiment, we also found that directly optimising $F_{\bm{\mathcal{T}}}(\mathbf{t})$ can cause instability. \[prop\_sampling\] The maximum of $\mathcal{C}_{\bm{\mathcal{T}}}(\bm{\mathcal{X}}, \bm{\mathcal{Y}})$ is reached when $F_{\bm{\mathcal{T}}}(\mathbf{t})$ attains a mass point at $\mathbf{t}^*$, where $\mathbf{t}^* = \arg\max_{\mathbf{t}}c(\mathbf{t})$. The minimum of $\mathcal{C}_{\bm{\mathcal{T}}}(\bm{\mathcal{X}}, \bm{\mathcal{Y}})$ is reached when $F_{\bm{\mathcal{T}}}(\mathbf{t})$ attains a mass point at $\mathbf{0}$. In the way of addressing this ill-posed optimisation on $F_{\bm{\mathcal{T}}}(\mathbf{t})$, we can impose some constraints on $F_{\bm{\mathcal{T}}}(\mathbf{t})$, for example, by assuming some parametric distributions. On the other hand, we may also be concerned that the constraints on $F_{\bm{\mathcal{T}}}(\mathbf{t})$ can impede the ability of $\mathcal{C}_{\bm{\mathcal{T}}}(\bm{\mathcal{X}}, \bm{\mathcal{Y}})$ as a metric to distinguish $\bm{\mathcal{X}}$ from $\bm{\mathcal{Y}}$. Lemma \[lemma\_sampling\] provides an efficient and feasible way of choosing $F_{\bm{\mathcal{T}}}(\mathbf{t})$. \[lemma\_sampling\] If $\bm{\mathcal{X}}$ and $\bm{\mathcal{Y}}$ are supported on a finite interval $[-1, 1]$, $\mathcal{C}_{\bm{\mathcal{T}}}(\bm{\mathcal{X}}, \bm{\mathcal{Y}})$ in is still a distance metric for distinguishing $\bm{\mathcal{X}}$ from $\bm{\mathcal{Y}}$ for any $F_{\bm{\mathcal{T}}}(\mathbf{t})$ that samples $\mathbf{t}$ within a small ball around $\mathbf{0}$. As will be introduced in the next section, we employ $\mathcal{C}_{\bm{\mathcal{T}}}(\bm{\mathcal{X}}, \bm{\mathcal{Y}})$ as the loss to compare two distributions from the *critic*. By employing bounded activation functions (tanh, sigmoid, etc.), the requirement of Lemma \[lemma\_sampling\] is automatically satisfied. Therefore, instead of searching within all the real distribution spaces, the choices of $F_{\bm{\mathcal{T}}}(\mathbf{t})$ can be safely restricted to some zero-mean distributions, e.g., the Gaussian distribution. Furthermore, compared to the fixed Gaussian distribution, it is preferable that, whilst avoiding the ill-posed optimum, that $F_{\bm{\mathcal{T}}}(\mathbf{t})$ could be optimised to better accommodate the difference between two distributions. In this paper, we choose $F_{\bm{\mathcal{T}}}(\mathbf{t})$ as the cdf of a broad class of distributions called the *scale mixture of normals*, in the form of $$\label{eq_scalemixture} p_{\bm{\mathcal{T}}}(\mathbf{t}) = \int_{\mathbf{\Sigma}} p_{\bm{\mathcal{N}}}(\mathbf{t}|\mathbf{0}, \mathbf{\Sigma})p_{\mathbf{\Sigma}}(\mathbf{\Sigma})d\mathbf{\Sigma},$$ where $p_{\bm{\mathcal{T}}}(\mathbf{t})$ is the pdf of $F_{\bm{\mathcal{T}}}(\mathbf{t})$, while $p_{\bm{\mathcal{N}}}(\mathbf{t}|\mathbf{0}, \mathbf{\Sigma})$ denotes the zero-mean Gaussian distribution with the covariance given by $\mathbf{\Sigma}$, and $p_{\mathbf{\Sigma}}(\mathbf{\Sigma})$ denotes distributions of $\mathbf{\Sigma}$. It needs to be pointed out that the *scale mixture of normals* constitutes a large portion of the elliptical distributions and includes many important distributions (e.g., the Gaussian, Cauchy, Student-*t*, hyperbolic distributions [@andrews1974scale]) by choosing different $p_{\mathbf{\Sigma}}(\mathbf{\Sigma})$. Therefore, instead of directly optimising $F_{\bm{\mathcal{T}}}(\mathbf{t})$, which leads to ill-posed solutions, we alternatively optimise the neural net to output the samples of $p_{\mathbf{\Sigma}}(\mathbf{\Sigma})$. By using the affine transformation (or the re-parametrisation trick), we are able to propagate back the gradients. Reciprocal Adversarial Learning =============================== Characteristic Function Loss in RCF-GAN --------------------------------------- Although a valid and complete metric for measuring any forms of distributions, the CF loss in works more efficiently and effectively in the embedded domain, than the data domain. To this end, we then express our RCF-GAN in the IPM-GAN format as $$\label{eq_IPM_ChF} d(\mathcal{P}_d, \mathcal{P}_g) = \sup_{\mathcal{T}, f\in\mathcal{F} }\mathcal{C}_{\bm{\mathcal{T}}}(f(\overline{\bm{\mathcal{X}}}), f(\overline{\bm{\mathcal{Y}}})), ~~\overline{\bm{\mathcal{X}}}\sim\mathcal{P}_d~\mathrm{and}~\overline{\bm{\mathcal{Y}}}\sim\mathcal{P}_g.$$ Lemma \[lemma\_differential\] below shows that this metric is well-defined for neural net training. \[lemma\_differential\] Given $f(\cdot)$ to be any differentiable function, the metric $\mathcal{C}_{\bm{\mathcal{T}}}(f(\overline{\bm{\mathcal{X}}}), f(\overline{\bm{\mathcal{Y}}}))$ in is bounded and differentiable almost everywhere. Because $\mathcal{C}_{\bm{\mathcal{T}}}(f(\overline{\bm{\mathcal{X}}}), f(\overline{\bm{\mathcal{Y}}}))$ is bounded by construction, it relaxes the requirements on the *critic* $f\in\mathcal{F}$. Otherwise, we may need to bound $\mathcal{F}$ to ensure the existence of the supremum [@mroueh2017fisher]. Furthermore, having proved that $\mathcal{C}_{\bm{\mathcal{T}}}(\bm{\mathcal{X}}, \bm{\mathcal{Y}}) \!=\! 0 \Leftrightarrow \bm{\mathcal{X}}\!=^d\!\bm{\mathcal{Y}}$, we also need to prove the equivalence between $\mathcal{C}_{\bm{\mathcal{T}}}(f(\overline{\bm{\mathcal{X}}}), f(\overline{\bm{\mathcal{Y}}})) \!=\! 0$ and $\overline{\bm{\mathcal{X}}}\!=^d\!\overline{\bm{\mathcal{Y}}}$, to ensure that our RCF-GAN correctly learns the real distribution in the data domain. This is provided in Lemma \[lemma\_injection\]. \[lemma\_injection\] Denote the distribution mapping by $\overline{\bm{\mathcal{Y}}}=^dg(\bm{\mathcal{Z}})$. Given two functions $f(\cdot)$ and $g(\cdot)$ that are reciprocal on the supports of $\overline{\bm{\mathcal{Y}}}$ and $\bm{\mathcal{Z}}$, that is, $\mathbb{E}_{\bm{\mathcal{Z}}}[||\mathbf{z} - f(g(\mathbf{z}))||^2_2]=0$, we also have $\mathbb{E}_{\overline{\bm{\mathcal{Y}}}}[||\overline{\mathbf{y}} - g(f(\overline{\mathbf{y}}))||^2_2]=0$. More importantly, this yields the following equivalences: $\mathcal{C}_{\bm{\mathcal{T}}}(f(\overline{\bm{\mathcal{X}}}), f(\overline{\bm{\mathcal{Y}}}))=0$ $\Leftrightarrow$ $\mathcal{C}_{\bm{\mathcal{T}}}(\overline{\bm{\mathcal{X}}}, \overline{\bm{\mathcal{Y}}})=0$ $\Leftrightarrow$ $\mathcal{C}_{\bm{\mathcal{T}}}(f(\overline{\bm{\mathcal{Y}}}), \bm{\mathcal{Z}})=0$ and $\mathcal{C}_{\bm{\mathcal{T}}}(f(\overline{\bm{\mathcal{X}}}), \bm{\mathcal{Z}})=0$. Putting Everything Together: Latent Space Matching -------------------------------------------------- Lemma \[lemma\_injection\] indicates that instead of being regarded as components of some IPMs (e.g., the W-GAN) to be optimised with strict restrictions, the *critic* can be basically regarded as feature mapping because in the embedded domain the CF loss is a valid distance metric of distributions. The *critic* can then be relaxed to satisfy the reciprocal property. Therefore, we incorporate the auto-encoder in only two modules by interchangeably treating the *critic* as the encoder and the generator as the decoder. Besides the computational ease, the structure of our RCF-GAN benefits both from the GAN and auto-encoder, as a way of unifying them. As an auto-encoder, the RCF-GAN enables to solely compare reconstructions in a meaningful embedded manifold instead of the data domain. When regarded as a GAN, the auto-encoder way theoretically and practically indicates the convergence; it also stabilises the training by pushing the embedded distributions to the Gaussian noise (i.e., $\bm{\mathcal{Z}}$). In practice, in Lemma \[lemma\_injection\], we regard $f(\cdot)$ as the *critic* and $g(\cdot)$ as the generator. For the *t*-net, we denote it as $h(\cdot)$ and assume the covariance matrix of its output is diagonal (we thus represent it as $\bm{\sigma}$), which is reasonable as in the embedded domain the multiple dimensions tend to be uncorrelated [@kingma2013auto]. Moreover, the goal of the *critic* is to maximise together with minimising an MSE loss to ensure the reciprocal property. Furthermore, to help our RCF-GAN in quickly mapping the real data distribution to the support of $\bm{\mathcal{Z}}$, we further re-design the *critic* loss by finding an anchor as $\mathcal{C}_{\bm{\mathcal{T}}}(f(\overline{\bm{\mathcal{Y}}}), \bm{\mathcal{Z}}) - \mathcal{C}_{\bm{\mathcal{T}}}(f(\overline{\bm{\mathcal{X}}}), \bm{\mathcal{Z}})$. On the other hand, the generator is trained by minimising as usual. The pseudo-code for our RCF-GAN is provided in Algorithm \[alg\]. \[alg\] Experimental Results ==================== In this section, our RCF-GAN is evaluated in terms of both random generations and reconstructions. We shall also point out that in this section, we report the results under the DCGAN structure [@radford2015unsupervised] for a fair comparison with other GANs, however, our RCF-GAN may achieve even better performances by using, for example, residual blocks, a deeper *t*-net, tailored $l_r$ in two scales, different weight $\alpha$ in , larger batch sizes or advanced techniques such as the spectral norm [@miyato2018spectral]. We report some of those improvements in the supplementary material, *including the achieved state-of-the-art results by using the residual nets*. Our code will be released at <https://github.com/ShengxiLi/rcf_gan>. **Datasets:** Three widely applied benchmark datasets were employed in our evaluation: CelebA (faces of celebrity) [@liu2015deep], CIFAR10 [@krizhevsky2009learning] and LSUN Bedroom (LSUN\_B) [@yu2015lsun]. The images of the CelebA and LSUN\_B were central cropped to the size $64\times64$, whist the image size of the CIFAR10 was $32\times32$. Furthermore, when evaluating the reconstruction, the test sets of the CIFAR10 and LSUN\_B were employed, from which the samples were not used in the training. **Baselines:** As our work is mainly related to the IPM-GANs, we compared our RCF-GAN with the W-GAN [@arjovsky2017wasserstein], W-GAN with gradient penalty (W-GAN-GP) [@gulrajani2017improved] and MMD-GAN [@li2017mmd; @binkowski2018demystifying]. For image reconstruction, we compared our RCF-GAN with the recent adversarial generator-encoder (AGE) work [@ulyanov2018takes], which empirically performs better than the adversarially learned inference (ALI) [@dumoulin2016adversarially]. **Metrics:** We employed the Fréchet inception distance (FID) [@heusel2017gans], which is basically the Wasserstein distance between two Gaussian distributions, together with the kernel inception distance (KID) that arises from the MMD metric [@binkowski2018demystifying]. **Net structure and technical details:** For a fair comparisons, all the reported results were compared under the batch sizes of 64 (i.e., $b_d\!=\!b_g\!=\!b_t\!=\!b_\sigma\!=\!64$) and also based on the network structure being the same as that in the DCGAN [@radford2015unsupervised], without any extra layers. In our RCF-GAN, we set $\lambda$ to linearly space from $0.6$ to $1$, which is reasonable as the role of *critic* changes from a discriminator to an encoder along the training procedure. Moreover, we set all variances of Gaussian noise to $1$, except for the input noise of the generator that was $0.3$, because the reciprocal loss has to be minimised given the fact that the output of the *critic* is restricted to $[-1,1]$. Furthermore, we do not require the Lipschitz constraint, which allows for a larger learning rate ($l_r\!=\!0.0002$ for both nets). Moreover, for the CIFAR10 and LSUN\_B datasets, the dimension of the embedded domain was set to $128$ and for the CelebA dataset the dimension was $64$. The optional *t*-net, if used, was a small three layer fully connected net, with the dimension of each layer being the same as the embedded dimension. **Image generation:** The generated images from random Gaussian noise are shown in Fig. \[figure\_genimgs\]. Observe that by using our proposed CF loss in the RCF-GAN, the generated images are clear and close to the real images. We further provide the FID and KID scores in Table \[table\_kidfid\]. This table shows that our RCF-GAN consistently achieved the best performances across the three datasets. More importantly, employing the *t*-net in outputting optimal $F_{\bm{\mathcal{T}}}(\mathbf{t})$ is beneficial, where further improvements are achieved. We also found that using the *t*-net resulted in better reconstructed images. More importantly, compared to a fluctuated generator loss that is caused by the adversarial module in GANs, we take the advantages of the auto-encoder structure in utilising the reciprocal loss (i.e., $\mathbb{E}_{\bm{\mathcal{Z}}}[||\mathbf{z} - f(g(\mathbf{z}))||^2_2]$ indicates the reciprocal, and the distance between the fake embedded and the Gaussian distributions), together with the distance between the embedded real distribution $f(\overline{\bm{\mathcal{X}}})$ and the Gaussian distribution $\bm{\mathcal{Z}}$ (i.e., $\mathcal{C}_{\bm{\mathcal{T}}}(f(\overline{\bm{\mathcal{X}}}), \bm{\mathcal{Z}})$) to better indicate the convergence, which are shown in Figure \[figure\_genimgs\]. The reciprocal loss basically measures the convergence on reconstructions, whereas the real image embedding distance $\mathcal{C}_{\bm{\mathcal{T}}}(f(\overline{\bm{\mathcal{X}}}), \bm{\mathcal{Z}})$ indicates the performance on generating images. **Image reconstruction:** The images reconstructed from the original images are shown in Fig. \[figure\_conimgs\]. Because our RCF-GAN only matches the distributions in the embedded domain, the reconstructed images are thus clear and semantically meaningful, resulting in a meaningful interpolation in Fig. \[figure\_conimgs\]. This is beneficial because rather than randomly generating real images, our RCF-GAN is able to bi-directionally reconstruct and interpolate real images. In contrast, although MMD-GANs employ a third module to achieve an auto-encoder, the decoded images are severely blurred. In the last three rows of Table \[table\_kidfid\], we can also see that the images reconstructed by our RCF-GAN are superior to those from the AGE. More importantly, the images from our RCF-GAN are consistently superior, whilst the quality of the reconstructed images in the AGE is much inferior to its random generated images. This also indicates the effectiveness of the unified structure of our RCF-GAN. Conclusion ========== We have introduced an efficient generative adversarial net (GAN) structure that seamlessly combines the IPM-GANs and auto-encoders. In this way, the reciprocal in our RCF-GAN ensures the equivalence between the embedded and data domains, whereas in the embedded domain the comparison on two distributions is strongly supported by the proposed powerful characteristic function (CF) loss, together with the physically meaningful phase and amplitude information, and an efficient sampling strategy. The reciprocal has been shown to also stabilise the convergence of the adversarial learning in our RCF-GAN, and at the same time to benefit from meaningful comparisons in the embedded domain. Consequently, the experimental results have demonstrated the superior performances of our RCF-GAN in both generating images and reconstructing images. Broader Impact ============== A combination of the auto-encoder and GANs has been extensively studied, and has been shown to achieve a broader data generation and reconstruction. The RCF-GAN proposed in this paper provides a neat and new structure in the combination. The studies of GANs and those design on probabilistic auto-encoders basically start from different perspectives because the former serves for the generation, or it “decodes” from random noise, whilst the latter, as its name implies, focuses on encoding to summarise information. Although there are extensive attempts on combining those two structures, they typically embed one into the other as components such as by using an auto-encoder as a discriminator in GANs or using an adversarial idea in an auto-encoder. This paper provides a way of equally treating the two structures; the proposed structure, which contains only two modules, can be regarded both as an “encoder-decoder” and “discriminator-generator”. The proposed combination benefits both, that is, it equips an auto-encoder the ability to meaningfully encode via matching in the embedded domain, whilst ensuring the convergence of the adversarial as a GAN. Moreover, instead of being a component to measure the distance as in the W-GAN, regarding the *critic* as an independent feature mapping module with a sufficient distance metric is beneficial to allow learning in the embedded domain for any types of feature extraction models, such as the deep canonical correlation analysis net and graph auto encoder. A large amount of unsupervised learning models, then, can be connected and improved with the adversarial learning. Another potential benefit of our work is to bring the general concept of the characteristic function (CF) into practice, by providing efficient sampling methods. The CF has been previously studied as a powerful tool in theoretical probabilistic analysis, while its practical applications have been limited due to complex functional forms. We should also highlight the physical meaning of the CF components introduced in this paper. It is a well known experimental phenomenon that the phase of discrete Fourier transform of images captures the saliency information, which motivates a large volume of works in saliency detection. This paper gives a probabilistic explanation to this, paving the way for future work to embark upon this intrinsic relationship. Additional and State-of-the-art Results ======================================= We evaluated the effectiveness and robustness of our RCF-GAN in two additional net structures, i.e., advanced DCGAN (adv-DCGAN) and residual net (ResNet) structures. In the adv-DCGAN structure, we increased the size of the generator whilst keeping the discriminator the same as that in the original DCGAN. For the ResNet structure, we adopted the same structure proposed in [@gulrajani2017improved]. We illustrate the difference between the generators of the adv-DCGAN and the original DCGAN in the following. [ ]{} Moreover, for the results on the DCGAN structure (reported in the paper), we trained our RCF-GAN for 150K iterations for both the CelebA and the CIFAR10 datasets, whereby 250K iterations were set to the LSUN\_B dataset. However, in the two additional structures (reported in the supplementary material), we set the maximum number of iterations to 100K for all the datasets, and also set the regulator fixed to $\lambda=0.8$ for simplicity. Although increasing the number of iterations and using linearly spaced $\lambda$ can gain further improvements, the achieved results are already the state-of-the-art, which is enough to validate the superior performances in bi-directionally generating clear images of our RCF-GAN. We report the overall results by extending the results of the paper in Table \[table\_kidfid\_ext\]. We can easily see that under the adv-DCGAN structure, our RCF-GAN consistently and stably achieved improvements by using a deeper generator. However, because the structure of the *critic* in the adv-DCGAN is the same as that in the DCGAN, the reconstruction quality of our RCF-GAN under the adv-DCGAN structure is comparable to that under the DCGAN structure. Furthermore, as proposed in [@gulrajani2017improved], the ResNet structure has been utilised in training the LSUN\_B dataset. We thus report in Table \[table\_kidfid2\] the comparisons under the ResNet structure, which is relatively larger in both the sizes of the generator and the *critic*, compared to the adv-DCGAN and DCGAN structures. In the following, we show the subjective results of the generation and reconstruction by our RCF-GAN. Proofs ====== ***Proof of Lemma \[lemma\_chfloss\_distance\]:*** We here prove the non-negativity, symmetry and triangle properties that are required as a valid distance metric in the following. **Non-negativity:** Based on the definition of $\mathcal{C}_{\bm{\mathcal{T}}}(\bm{\mathcal{X}}, \bm{\mathcal{Y}})$, $$\mathcal{C}_{\bm{\mathcal{T}}}(\bm{\mathcal{X}}, \bm{\mathcal{Y}}) = \int_\mathbf{t} \big(\underbrace{({\Phi}_{\bm{\mathcal{X}}}(\mathbf{t}) - {\Phi}_{\bm{\mathcal{Y}}}(\mathbf{t}))({\Phi}^*_{\bm{\mathcal{X}}}(\mathbf{t}) - {\Phi}^*_{\bm{\mathcal{Y}}}(\mathbf{t}))}_{c(\mathbf{t})}\big)^{\nicefrac{1}{2}}dF_{\bm{\mathcal{T}}}(\mathbf{t}),$$ the term $\mathcal{C}_{\bm{\mathcal{T}}}(\bm{\mathcal{X}}, \bm{\mathcal{Y}})$ is non-negative because $c(\mathbf{t})\geq 0$ for all $\mathbf{t}$. We next prove when the equality holds. - $\bm{\mathcal{X}}=^d\bm{\mathcal{Y}} \rightarrow \mathcal{C}_{\bm{\mathcal{T}}}(\bm{\mathcal{X}}, \bm{\mathcal{Y}})=0$: This is evident because $\Phi_{\bm{\mathcal{X}}}(\mathbf{t})=\Phi_{\bm{\mathcal{Y}}}(\mathbf{t})$ for all $\mathbf{t}$. - $\bm{\mathcal{X}}=^d\bm{\mathcal{Y}} \leftarrow \mathcal{C}_{\bm{\mathcal{T}}}(\bm{\mathcal{X}}, \bm{\mathcal{Y}})=0$: Given that the support of $\bm{\mathcal{T}}$ is $\mathbb{R}^m$, $\int_\mathbf{t}\sqrt{c(\mathbf{t})}dF_{\bm{\mathcal{T}}}(\mathbf{t})=0$ exists if and only if $c(\mathbf{t})=0$ everywhere. Therefore, $\Phi_{\bm{\mathcal{X}}}(\mathbf{t})=\Phi_{\bm{\mathcal{Y}}}(\mathbf{t})$ for all $\mathbf{t}\in\mathbb{R}^m$. According to the *Uniqueness Theorem* of the CF, we have $\bm{\mathcal{X}}=^d\bm{\mathcal{Y}}$. Therefore, we have that $\mathcal{C}_{\bm{\mathcal{T}}}(\bm{\mathcal{X}}, \bm{\mathcal{Y}})\geq 0$, and the equality holds if and only if $\bm{\mathcal{X}}=^d\bm{\mathcal{Y}}$. **Symmetry:** This is obvious for the symmetry of $c(\mathbf{t})$, thus yielding $\mathcal{C}_{\bm{\mathcal{T}}}(\bm{\mathcal{X}}, \bm{\mathcal{Y}}) = \mathcal{C}_{\bm{\mathcal{T}}}(\bm{\mathcal{Y}}, \bm{\mathcal{X}})$. **Triangle:** Because the CFs $\Phi_{\bm{\mathcal{X}}}(\mathbf{t})$ and $\Phi_{\bm{\mathcal{X}}}(\mathbf{t})$ are the elements of the normed vector space, we have the following inequality (also known as the Minkowski inequality), $$\begin{aligned} \int_{\mathbf{t}}|\Phi_{\bm{\mathcal{X}}}(\mathbf{t}) - \Phi_{\bm{\mathcal{Z}}}(\mathbf{t}) &+ \Phi_{\bm{\mathcal{Z}}}(\mathbf{t}) - \Phi_{\bm{\mathcal{Y}}}(\mathbf{t})|dF_{\bm{\mathcal{T}}}(\mathbf{t}) \\ &\leq \int_{\mathbf{t}}|\Phi_{\bm{\mathcal{X}}}(\mathbf{t}) - \Phi_{\bm{\mathcal{Z}}}(\mathbf{t})|dF_{\bm{\mathcal{T}}}(\mathbf{t}) + \int_{\mathbf{t}}|\Phi_{\bm{\mathcal{Z}}}(\mathbf{t}) - \Phi_{\bm{\mathcal{Y}}}(\mathbf{t})|dF_{\bm{\mathcal{T}}}(\mathbf{t}). \end{aligned}$$ Therefore, the triangle property of $\mathcal{C}_{\bm{\mathcal{T}}}(\bm{\mathcal{X}}, \bm{\mathcal{Y}})$ follows as $$\mathcal{C}_{\bm{\mathcal{T}}}(\bm{\mathcal{X}}, \bm{\mathcal{Y}}) \leq \mathcal{C}_{\bm{\mathcal{T}}}(\bm{\mathcal{X}}, \bm{\mathcal{Z}}) + \mathcal{C}_{\bm{\mathcal{T}}}(\bm{\mathcal{Z}}, \bm{\mathcal{Y}}).$$ Therefore, $\mathcal{C}_{\bm{\mathcal{T}}}(\bm{\mathcal{X}}, \bm{\mathcal{Y}})$ is a valid distance metric in measuring discrepancies between two random variables $\bm{\mathcal{X}}$ and $\bm{\mathcal{Y}}$. This completes the proof. ***Proof of Lemma \[lemma\_sampling\]:*** The proof of the triangle and symmetry properties is the same as those in Lemma \[lemma\_chfloss\_distance\]. The non-negativity is also evident and the same as that in Lemma \[lemma\_chfloss\_distance\] but the equality holds for different conditions. We provide its proof in the following. Before introducing our proof, we first quote Theorem 3 from Essen [@esseen1945fourier]. It is the fact that only requiring $\Phi_{{\mathcal{X}}}(\mathbf{t})=\Phi_{{\mathcal{Y}}}(\mathbf{t})$ in an interval around $\mathbf{0}$ does not ensure the equivalence between two distributions without any other constraints, given the counterexample provided in [@esseen1945fourier] as well. This equivalence even cannot be ensured when all the moments are matched. The third condition, intuitively, guarantees this equivalence by restricting that the moment does not increase “extremely” fast when $k\rightarrow \infty$. In Lemma \[lemma\_sampling\] of this paper, we bound $\bm{\mathcal{X}}$ and $\bm{\mathcal{Y}}$ by $[-1, 1]$, thus having $|\beta_k| \leq 1 < \infty$ and $\nicefrac{1}{\beta_{2k}^{\nicefrac{1}{2k}}} \geq 1$ so that $\sum_{k=1}^\infty \nicefrac{1}{\beta_{2k}^{\nicefrac{1}{2k}}}$ diverges. In this case, according to Theorem 3, we have $\Phi_{{\mathcal{X}}}(\mathbf{t})=\Phi_{{\mathcal{Y}}}(\mathbf{t})$ when $\mathbf{t}$ samples around $\mathbf{0} \rightarrow \bm{\mathcal{X}}=^d\bm{\mathcal{Y}}$. Conversely, it is obvious that $\bm{\mathcal{X}}=^d\bm{\mathcal{Y}} \rightarrow \Phi_{{\mathcal{X}}}(\mathbf{t})=\Phi_{{\mathcal{Y}}}(\mathbf{t})$ for all $\mathbf{t}$. Therefore, as for bounded $\bm{\mathcal{X}}$ and $\bm{\mathcal{Y}}$, sampling around $\mathbf{0}$ is sufficient to ensure the symmetry, triangle, non-negativity (together with the uniqueness when the equality holds) properties of $\mathcal{C}_{\bm{\mathcal{T}}}(\bm{\mathcal{X}}, \bm{\mathcal{Y}})$. This completes the proof. ***Proof of Lemma \[lemma\_differential\]:*** We first show the boundedness of $\mathcal{C}_{\bm{\mathcal{T}}}(\bm{\mathcal{X}},\bm{\mathcal{Y}})$ by observing $$\begin{aligned} 0\leq\mathcal{C}_{\bm{\mathcal{T}}}(\bm{\mathcal{X}},\bm{\mathcal{Y}}) &= \int_\mathbf{t}|\Phi_{\bm{\mathcal{X}}}(t)-\Phi_{\bm{\mathcal{Y}}}(t)|dF_{\bm{\mathcal{T}}}(\mathbf{t}) \\ &\leq \int_\mathbf{t}|\Phi_{\bm{\mathcal{X}}}(t)|dF_{\bm{\mathcal{T}}}(\mathbf{t})+\int_{\mathbf{t}}|\Phi_{\bm{\mathcal{Y}}}(t)|dF_{\bm{\mathcal{T}}}(\mathbf{t}) \leq 1+1 = 2, \end{aligned}$$ where the second inequality is obtained via the Minkowski inequality and the third one by the fact that the maximal modulus of the CF is $1$. It needs to be pointed out that this property is important and advantageous because in this way our cost is bounded automatically. Otherwise, we may need to bound $f\in\mathcal{F}$ to ensure an existence of the supremum of some IPMs (such as the dual form of the Wasserstein distance used in the W-GAN). To prove the differentiable property, we first expand $c(\mathbf{t})$ in $\mathcal{C}_{\bm{\mathcal{T}}}(\bm{\mathcal{X}},\bm{\mathcal{Y}}) = \int_\mathbf{t}\sqrt{c(\mathbf{t})}dF_{\bm{\mathcal{T}}}(\mathbf{t})$ as $$c(\mathbf{t}) = (\mathrm{Re}\{\Phi_{{\bm{\mathcal{X}}}}(\mathbf{t})\} -\mathrm{Re}\{\Phi_{\bm{{\mathcal{Y}}}}(\mathbf{t})\})^2 - (\mathrm{Im}\{\Phi_{\bm{{\mathcal{X}}}}(\mathbf{t})\}-\mathrm{Im}\{\Phi_{\bm{{\mathcal{Y}}}}(\mathbf{t})\})^2,$$ where $\mathrm{Re}\{\Phi_{{\bm{\mathcal{X}}}}(\mathbf{t})\}=\mathbb{E}_{\bm{\mathcal{X}}}[\cos(\mathbf{t}^T\mathbf{x})]$ denotes the real part of the CF and $\mathrm{Im}\{\Phi_{{\bm{\mathcal{X}}}}(\mathbf{t})\}=\mathbb{E}_{\bm{\mathcal{X}}}[\sin(\mathbf{t}^T\mathbf{x})]$ for its imaginary part. Therefore, by regarding $c(\mathbf{t})$ as a mapping $\mathbb{R}^m\rightarrow\mathbb{R}$, it is differentiable almost everywhere[^3]. Due to the fact that the composition of differentiable functions is also differentiable, we conclude that $\mathcal{C}_{\bm{\mathcal{T}}}(f(\overline{\bm{\mathcal{X}}}),f(\overline{\bm{\mathcal{Y}}}))$ is differentiable and continuous almost everywhere when $f(\cdot)$ is differentiable (e.g., the neural net). This completes the proof. ***Proof of Lemma \[lemma\_injection\]:*** Because $\mathbb{E}_{\bm{\mathcal{Z}}}[||\mathbf{z} - f(g(\mathbf{z}))||_2^2]=0$ and $||\mathbf{z} - f(g(\mathbf{z}))||_2^2\geq0$, we have $\mathbf{z} = f(g(\mathbf{z}))$ for any $\mathbf{z}$ and $g(\mathbf{z})$ under the supports of $\bm{\mathcal{Z}}$ and $\overline{\bm{\mathcal{Y}}}$, respectively. We can obtain $g(\mathbf{z}) = g(f(g(\mathbf{z})))$ under the supports of $\bm{\mathcal{Z}}$ and $\overline{\bm{\mathcal{Y}}}$ as well; given that $\overline{\mathbf{y}}=g(\mathbf{z})$ by the definition, this results in $\overline{\mathbf{y}} = g(f(\overline{\mathbf{y}}))$. Then, we have $\mathbb{E}_{\overline{\bm{\mathcal{Y}}}}[||\overline{\mathbf{y}} - g(f(\overline{\mathbf{y}}))||^2_2]=0$. Therefore, the function $g(\cdot)$ is a unique inverse of the function $f(\cdot)$, and vice versa, which also indicates that the two functions are bijective. The bijection of the function $f(\cdot)$ possesses many desirable properties between the domains of $\overline{\bm{\mathcal{Y}}}$ and $\bm{\mathcal{Z}}$, thus ensuring the equivalences between their CFs. Specifically, without loss of generality, we assume $\mathcal{C}_{\bm{\mathcal{T}}}(\overline{\bm{\mathcal{X}}}, \overline{\bm{\mathcal{Y}}})=0$, which means $$\int_{\overline{\mathbf{x}}}e^{j\overline{\mathbf{t}}^T\overline{\mathbf{x}}}dF_{\overline{\bm{\mathcal{X}}}}(\overline{\mathbf{x}}) = \int_{\overline{\mathbf{y}}}e^{j\overline{\mathbf{t}}^T\overline{\mathbf{y}}}dF_{\overline{\bm{\mathcal{Y}}}}(\overline{\mathbf{y}}),~~~~\mathrm{for~all~} \overline{\mathbf{t}}.$$ Then, given the bijection $f(\cdot)$ by $\bm{\mathcal{X}}=f(\overline{\bm{\mathcal{X}}})$ and $\bm{\mathcal{Y}}=f(\overline{\bm{\mathcal{Y}}})$, we can obtain that $\mathbf{x}=f(\overline{\mathbf{x}})=f(\overline{\mathbf{y}})=\mathbf{y}\Leftrightarrow\overline{\mathbf{x}}=\overline{\mathbf{y}}$, for any realisations $\mathbf{x}$ and $\mathbf{y}$ from $\bm{\mathcal{X}}$ and $\bm{\mathcal{Y}}$. We then have the following equivalence between the CFs of $\bm{\mathcal{X}}=f(\overline{\bm{\mathcal{X}}})$ and $\bm{\mathcal{Y}}=f(\overline{\bm{\mathcal{Y}}})$, $$\begin{aligned} &\int_{\overline{\mathbf{x}}}e^{j\overline{\mathbf{t}}^T\overline{\mathbf{x}}}dF_{\overline{\bm{\mathcal{X}}}}(\overline{\mathbf{x}}) = \int_{\overline{\mathbf{y}}}e^{j\overline{\mathbf{t}}^T\overline{\mathbf{y}}}dF_{\overline{\bm{\mathcal{Y}}}}(\overline{\mathbf{y}}),~~~~\mathrm{for~all~} \overline{\mathbf{t}}\\ &\Leftrightarrow \int_{\overline{\mathbf{x}}}e^{j{\mathbf{t}}^Tf(\overline{\mathbf{x}})}dF_{\overline{\bm{\mathcal{X}}}}(\overline{\mathbf{x}}) = \int_{\overline{\mathbf{y}}}e^{j{\mathbf{t}}^Tf(\overline{\mathbf{y}})}dF_{\overline{\bm{\mathcal{Y}}}}(\overline{\mathbf{y}}),~~~~\mathrm{for~all~} {\mathbf{t}} \\ &\Leftrightarrow \int_{{\mathbf{x}}}e^{j{\mathbf{t}}^T{\mathbf{x}}}dF_{{\bm{\mathcal{X}}}}({\mathbf{x}}) =\int_{{\mathbf{y}}}e^{j{\mathbf{t}}^T{\mathbf{y}}}dF_{{\bm{\mathcal{Y}}}}({\mathbf{y}}),~~~~\mathrm{for~all~} {\mathbf{t}}. \end{aligned}$$ Therefore, we have $\mathcal{C}_{\bm{\mathcal{T}}}(f(\overline{\bm{\mathcal{X}}}), f(\overline{\bm{\mathcal{Y}}}))=0$. Furthermore, we also have $f(\overline{\bm{\mathcal{Y}}})=^d\bm{\mathcal{Z}}$ due to $\mathbb{E}_{\bm{\mathcal{Z}}}[||\mathbf{z} - f(g(\mathbf{z}))||_2^2]=0$. Therefore, we have the following equivalences: $\mathcal{C}_{\bm{\mathcal{T}}}(\overline{\bm{\mathcal{X}}}, \overline{\bm{\mathcal{Y}}})=0\Leftrightarrow\mathcal{C}_{\bm{\mathcal{T}}}(f(\overline{\bm{\mathcal{X}}}), f(\overline{\bm{\mathcal{Y}}}))=0\Leftrightarrow\mathcal{C}_{\bm{\mathcal{T}}}(f(\overline{\bm{\mathcal{Y}}}), \bm{\mathcal{Z}})=0$ and $\mathcal{C}_{\bm{\mathcal{T}}}(f(\overline{\bm{\mathcal{X}}}), \bm{\mathcal{Z}})=0$. This completes the proof. Related Works ============= **IPM-GANs:** Instead of the naive weight clipping in the W-GAN [@arjovsky2017wasserstein], the gradient penalty W-GAN (W-GAN-GP) was proposed to mitigate the heavily constrained *critic* by penalising the gradient norm [@gulrajani2017improved], followed by a further elegant treatment of restricting the largest singular value of the net weights [@miyato2018spectral].It has been understood that although the *critic* cannot search within all satisfied Lipschitz functions [@arora2017generalization; @arora2017gans], the *critic* still performs as a way of transforming high dimensional but insufficiently supported data distributions into low dimensional yet broadly supported (simple) distributions in the embedded domain [@binkowski2018demystifying]. Comparing the embedded statistics, however, is much easier. For example, Cramer GAN compared the mean with an advanced $\mathcal{F}$ from the Cramer distance to correct the biased gradient [@bellemare2017cramer], whilst McGAN [@mroueh2017mcgan] explicitly compared the mean and the covariance in the embedded domain. Fisher GAN employed a scale-free Mahalanobis distance and thus data dependent $\mathcal{F}$ [@mroueh2017fisher], which is basically the Fisher-Rao distance in the embedded domain between two Gaussian distributions with the same covariance. The recent Sphere GAN further compared higher-order moments up to a specified order, and avoided the Lipschitz condition by projecting onto a spherical surface [@park2019sphere]. Moreover, in a non-parametric way, BE-GAN directly employed an auto-encoder as the *critic*, whereby the auto-encoder loss was compared through embedded distributions [@berthelot2017began]. The sliced Wasserstein distance has also been utilised in measuring the discrepancy in the embedded domain [@deshpande2018generative]. Another non-parametric metric was achieved by the kernel trick of the MMD-GAN [@li2017mmd; @binkowski2018demystifying], which assumed $\mathcal{F}$ as the reproducing kernel Hilbert space. However, as one of the most powerful ways of representing a distribution, the CF, is still to be fully explored. More importantly, our RCF-GAN fully compares the embedded distributions and also generalises the MMD-GAN by flexible sampling priors. **Auto-encoders in an adversarial way:** To address the smoothing artefact of the variational auto-encoder [@kingma2013auto], several works aim to incorporate the adversarial style in (variational) auto-encoders, in the hope of gaining clear images whilst maintaining the ability of reconstruction. These mostly consist of at least three modules, an encoder, a decoder, and an adversarial modules [@makhzani2015adversarial; @larsen2015autoencoding; @donahue2016adversarial; @brock2016neural; @che2016mode; @dumoulin2016adversarially]. To the best of our knowledge, the one exception, called the adversarial generator encoder (AGE) [@ulyanov2018takes], only incorporated two modules in adversarially training an auto-encoder under a max-min problem. The AGE still assumes the Gaussianity in the embedded distributions and only compares the mean and the diagonal covariance matrix; this is basically insufficient in identifying two distributions, in which the pixel domain loss was utilised supplementally. However, the initiative of this paper is fundamentally different from the AGE, and the auto-encoder in our RCF-GAN aims to achieve the theoretical guarantee of a bijection, with different losses proposed for the *critic* and the generator. Furthermore, other than the first- and second-order moments, our work fully compares the discrepancies in the embedded domain via CFs. Benefiting from the powerful non-parametric metric via the CFs, our RCF-GAN only adversarially learns distributions in the embedded domain, that is, on a semantically meaningful manifold, without the need of any operation on the data domain. [^1]: To our best knowledge, the only exception is the AGE [@ulyanov2018takes], which adopts two modules in an auto-encoder under a max-min problem and different losses. Please see the *Related Works* for the details. [^2]: We notice a very recent independent work [@ansari2020characteristic] which also employs the CF as a replacement by using the same structure of MMD-GANs. Our RCF-GAN, however, is substantially different from that in [@ansari2020characteristic] in terms of the analysis on decomposing CFs, CF losses, theoretical guarantees, GAN structures, etc. Our achieved results are also superior to those in [@ansari2020characteristic], where all images were evaluated under the size of $32\times32$. [^3]: Please note that $c(\mathbf{t})$ is not necessarily complex differentiable because it does not satisfy the Cauchy-Riemann equations. It is the fact that any *nonconstant purely real-valued functions* are not complex differentiable because their Cauchy-Riemann equations cannot be satisfied. However, in our case of regarding it as mappings in the real domain, it is differentiable. Please refer to [@kreutz2009complex; @mandic2009complex] for more detail in the $\mathbb{CR}$ calculus. | High | [
0.6728971962616821,
31.5,
15.3125
] |
After a long break, ethnic clashes have returned to northern Kosovo. On July 25the crisis reopened, when special Kosovo police units (ROSU) surprised half-empty Serbian towns in northern Kosovo and rushed into the two administrative entry points into central Serbia - Brnjak and Jarinje. Kosovo’s government recently closed their southern border to products from Serbia, until Serbia recognizes the custom stamp of the Kosovo state. Rumours that the same action might follow on Kosovo’s northern border had not stopped many Serbs in the border region to leave their homes for a holiday. The remaining residents immediately reacted to the Kosovo police action by blocking the roads with trucks and erecting barricades of woods and sand. Kosovo Albanians managed to arrive in Brnjak and disarm the Kosovo Police Service (KPS) officers (ethnic Serbs) there. They were then caught in ‘a sandwich’ when they tried to do the same in Jarinje. The same night the Serbian government’s head negotiator with Pristina, Borislav Stefanović, accompanied by Kosovo Minister Goran Bogdanović arrived to meet with KFOR commander Erhard Bühler and to stay with the local community. In the meeting, Serbs requested that the special ROSU troops leave the north and return the situation to what it had been before their arrival, when the ethnic Serbian KPS officers and EULEX police were in charge. Kosovo’s central government was determined that ROSU stay and establish the same custom system as on the rest of Kosovo’s borders. The death of policeman Enver Zymberi In spite of positive signals to resolve the crisis, the marathon meeting appeared not to be the last. The following day, ROSU and local Serbs in the Zubin Potok area exchanged gunfire that resulted in the death of a member of the special police ROSU unit. Albanian sources reported that a sniper bullet to his head killed him while he was withdrawing from Brnjak. An eyewitness, who wished to remain anonymous, claimed that the Kosovo Albanian policeman had been killed in the exchange of automatic fire: “As was agreed, a ROSU unit was withdrawing. However, as soon as they left the area we control, a new ROSU unit started driving towards our barricades. We immediately rushed with cars towards them before they would reach the barricades because we wanted to stop them from entering the town. As we met on the road, they could not pass our cars; they stopped and stepped out with the guns. There were shots in the air, I do not know who shot first, and then we started shooting in the direction of each other. They broke the deal. They should have not come back.” After the continued talks between the Serbs and KFOR the same day, Stefanović said that all the details were agreed and the last troops of ROSU would finally pull out by 8am the following day. Back in Pristina the same evening, Kosovo’s Prime Minister Hashim Thaci made an extraordinary press statement to honour the killed officer and justify the action. Without any possibility for journalists to ask questions, Thaci declared: “Last night’s decision was not an easy one, but it was a just one and such an act was necessary. Finally, as a Prime Minister I have a moral, constitutional and legal duty to make such hard decisions.” In Kosovo, the killed officer Enver Zymberi was proclaimed a hero. Several Facebook pages in his honour were immediately launched after his death and rallied thousands of ethnic Albanian sympathisers. They shared strong national sentiments, condemning Serbs. An NGO activist, Valjdete Idrizi, from southern Mitrovica, proclaimed: “He was killed while protecting his state’s border, murdered on duty, murdered in an ambush by a sniper.” State of emergency Kosovo Interior Minister Bajram Rexhepi confirmed yesterday that ROSU had pulled out of North Kosovo, but stated that the ROSU mission was successfully accomplished after it established a custom border and created conditions for the work of an ethnically mixed border police. Although the crisis seemed resolved, the following day contradicting information quickly spread about ROSU units staying in the American KFOR’s base in Leposavić and that Croatian helicopters were transporting new ROSU units to Brnjak. Croatian sources later confirmed that someone shot at their helicopter transporting custom and border police. Although Bühler denied the presence of any ROSU unit in the north while addressing massed Kosovo Serbs, the panic and tensions increased when local authorities declared a crisis situation. Local news reported that Albanian customs had taken over Brnjak and that an American KFOR motorcade had been stopped at the Zvečan barricade, when travelling to Leposavić. The rumours spread that ROSU troops were transported by the American Humvee vehicles and helicopters to Jarinje. In the meantime, Stefanović insisted on the earlier agreed upon deal and requested the withdrawal of Kosovo Albanian custom officers and border police from the Brnjak checkpoint. Accompanied by Kosovo’s moderate Serbian bishop Teodosije, he tried to calm down the tensions among the local Serbs and finally persuaded them to let the Americans use the road, after having received insurances they were not transporting any ROSU troops but only insuring regular food and water supply for the military base in Leposavić. Barricades Barricades in Zubin Potok (Foto Tatjana Lazarević) People on the barricades have been sleepless for nights, angry, tense and sensitive to any rumour or change of the situation. They unwillingly speak to media and look highly mistrustful, refusing to reveal their names or to be photographed. A protester at the barricade in nearby village Zupče, in Zubin Potok municipality softly said: “You have to understand us. We are all the time outside in this hot weather. Most of all, we are angry with KFOR. They blame us, but the Kosovo Albanians provoked the crisis. We do not want an Albanian custom or border station here. It is our land and system. We defend; we do not attack.” A tall, wide-shouldered young man angrily shouted: “We have no future in their country. Look at the Serbs in the south. They are disappearing.” The Mayor of Zubin Potok, Slaviša Ristić was with his men. He harshly said: “They lied to us. Bühler told us yesterday that if we spot any ROSU troops, we should immediately let him know it, to arrest them, and today, they are transporting them to the checkpoint by helicopters.” Torching of the check-point The culmination was yet to come in the evening. All of a sudden, while a few Serbian TV stations had programs live from the quiet Jarinje gate, about 50 youngsters in hoods and with sunglasses suddenly emerged in front of the cameras, furiously torched all the facilities, bulldozed the area and disappeared. Until now, it remained unclear how they passed the barricades. EULEX armoured vehicles later sealed the main bridge in Mitrovica North. The violence was immediately followed by a strong and angry reaction by the Serbian president and police officials, along with the Serbian crisis-solving representative Stefanović. Some political officials issued contradictory statements on whether the violence was well organized or spontaneous. The Serbian police started heavy controls to prevent the possible arrival of troublemakers from central Serbia. Agitators from Pristina immediately accused Belgrade. Both sides agreed in labelling the perpetrators as extremists, hooligans and vandals. Once again, Stefanović and Bühler met late into the night. American KFOR today controls what remains of the Jarinje checkpoint and supports Brnjak, which still has Kosovo custom official and ethnic Albanian Kosovo border police. Stefanović and Bühlers are still talking today, the fourth day of the crisis. How a crisis escalates Fuel for the fire comes from the constant circulation of rumours and contradictory, unverified information. Zvezdan Djukanović, an Associated Press correspondent and photographer, said: “It was unbelievably difficult to work these days, due to the mass of misinformation and rumours we were getting.” On the other side, EULEX remains strikingly silent. In a sterile press statement, they stated: “It is important that the current situation is resolved in a calm and peaceful manner. And EULEX is coordinating closely with KFOR to do this and to maintain security.” They denied any involvement in the ROSU operation. After almost an artificial silence, northern Kosovo is back to its infamous unruliness. The tension is in the air. Helicopters hover. Protesters are back on the streets. A fearless and furious new generation of teenagers has been raised in a barbed-wire city, while watching ethnic clashes on their streets since their earliest childhood. The older ones know of the danger that northern Kosovo teeters at the edge of a wide-scale clash. All of them are adamantly fighting back against facing a similar scenario to that of Serbs in southern Kosovo. | Low | [
0.52127659574468,
30.625,
28.125
] |
Science academies of 13 nations urge G8 to tackle climate change Ahead of the G8 summit in Japan next month, the science academies of 13 nations, including the United States, urged the G8 nations as well as Brazil, China, India, Mexico, and South Africa (G8+5) to agree to cut world greenhouse-gas emissions in half by 2050. “We urge G8+5 leaders to make maximum efforts to carry this forward and commit to these emission reductions,” the academies said in a statement. “Progress in reducing global greenhouse-gas emission has been slow. … Key vulnerabilities include water resources, food supply, health, coastal settlements, and some ecosystems, particularly Arctic, tundra, alpine, and coral reef.” This is not the first time the academies have pleaded for meaningful climate action, and it likely won’t be the last time their counsel is ignored. At last year’s summit, despite similar urging from scientists, only six of the G8 nations came out in favor of pursuing the halving-emissions goal — the U.S. and Russia opposed it. | Mid | [
0.650537634408602,
30.25,
16.25
] |
Charles Harmon Director of University Relations, Sementha Mathews Manager of Public Information and Media Relations College of Education receives new name at dedication ceremony Valdosta State University's College of Education will be officially named the James L. and Dorothy H. Dewar College of Education, at a naming dedication ceremony, Tuesday, April 4, 2006, at 11:00 a.m., on the Baytree Road side of the Education Center. The new name reflects the university's appreciation of a $2 million gift from the Dewar family and Park Avenue Bank, founded by Dewar. The $2 million gift is one of the largest single gifts received by VSU and will be used to help implement projects that enhance the benefits provided by current programs and further the national reputation of the College. Limited parking is available in the Oak Street parking lot, and additional parking will be available in the Campus Recreation parking lot on Sustella Avenue. For more information, call the Office of Special Events at 333-7446. | Mid | [
0.633027522935779,
34.5,
20
] |
Channelized Rivers With the development of channelized rivers, our water systems have not been the same. Channelized rivers of the Los Angeles watershed impacted water in numerous ways. Breaking it down to the pros of building a watershed along our water resources, the channelized rivers mitigate chances of having rapid floods. In response to that, Los Angeles is able to build their city along rivers. With this, channelized rivers mainly benefit people. As oppose to building channelized rivers, the cons include the depletion of our local water resources. Los Angeles loses their water resources due to the water flowing down to the ports of Long Beach rather quickly. Additionally, the massive amounts of concrete include destruction of wildlife, collection of unsettled contaminated water, and rapid currents downstream. Determining the better option of channelized rivers or not, it is a matter of opinion. Channelized rivers contain the benefits for human development; however, channelized waters damage our ecosystem. Whether it’s channelized or not, we need water to survive. Like this: Related Post navigation 4 responses to “Channelized Rivers” Channelization impacts the ecosystems in which the water is being siphoned from. Although we do need water to survive, it is still our fault as to why we have an issue with making channelized rivers in the first place. Humans are a self centered breed that chooses to populate areas of visual pleasure. If planning were made with a more environmental or sustainable approach, water problems would not be as severe as they are now. It is the original planners who made the mistakes and now it is up to us to fix their problems. Yes, humans, need water, but don’t we need air, soil, vegetation to survive? Yes, we do, but many forget or unknown the importance of these others natural resources as well. At least lets do it for ourselfs and take care of them. Channelization, although impacts the environment in a negative way is something that we have to deal with. i feel that the best solution is to find alternative ways and or materials of channelization because whether we like it or not it will still occur. finding ways to mitigate water flows through vegetation or materials other than concrete that can be permeable are some ways that we can get the same result without having the same impact on the environment. instead of opposing the problem we should use it as a foundation for sustainable solution. Ultimately, these cement ditches must go if we are to become sustainable here in S. Calif. We cannot just “channel” away our water to the ocean, as the population and water needs rise. We need to remove the concrete, and build multiple lakes and settling ponds between the mountains and the ocean. | Mid | [
0.5900473933649291,
31.125,
21.625
] |
1. Field of the Invention The present invention relates to a wire bonding device and a wire bonding method in which the impact load applied on a bonding subject at the time of wire contact detection can be reduced and a highly precise and stable bonding can be achieved by a minute bonding ball diameter while enabling to achieve a high productivity through high-speed bonding. 2. Description of the Related Art Conventionally, it has been required for a wire bonding device to have a structure in which bonding arms are extended in order to avoid interference between the conveyor/fixing mechanism of the bonding subject. It is to satisfy the needs for expanding the field of bonding aiming at large-scaled bonding subjects and improving productivity. On the other hand, in order to cope with the narrowing intervals of the wire bonding portion of the bonding subject, it has been desired that the shape of the bonding portion be made minute with high precision. FIG. 5 shows a schematic model block diagram of an example of the wire bonding device of the related art. A wire bonding device of the related art comprises: a Z-axis base 101 mounted on an XY table (not shown) having a unit which drives in the X and Y directions crossing orthogonal to each other; a bonding arm 103 supported by the Z-axis base 101 via a shaft 102 to be capable of swinging; and a VCM (voice coil motor) 104 for driving the bonding arm 103 in the Z-axis direction. Provided in the bonding arm 103 are an ultrasonic horn 109 having a capillary 108 to which a wire 111 made of a gold wire is inserted and a cut clump 110 for cutting the wire 111. The bonding arm 103 has the length long enough to avoid the interference with the conveyor/fixing mechanism of the semiconductor integrated circuit (not shown) as the bonding subject. An outline of wire bonding performed by the above-described wire bonding device of the related art will be described below. First, after forming an initial ball in the tip of the wire 111 extended from the bottom end of the capillary 108, the bonding arm 103 is brought down by the VCM 104 until it comes to be in contact with a first bonding point of the semiconductor integrated circuit. At the time of contact detection when the initial ball comes to be in contact with the semiconductor integrated circuit, the bonding arm 103 is driven at low speed in order to reduce impact load applied on the semiconductor integrated circuit. Subsequently, ultrasonic vibrations generated by an ultrasonic oscillator (not shown) is transmitted to the initial ball via the ultrasonic horn 109 and the capillary 108 while applying a prescribed weight thereto so as to bond the wire 111 to the first bonding point. After completing bonding at the first bonding point, the capillary 108 is brought up through driving the bonding arm 103 by the VCM 104. Then, the XY table is moved horizontally to the position over a second bonding point and the wire 111 is pressed to be bonded to the second bonding point in the same manner as that in the first bonding point. Subsequently, the bonding arm 103 is brought up and the wire 111 above the capillary 108 is tightly held by the cut clamp 110 and pulled up. Thereby, the wire 111 is cut at the joint of the second bonding point. Through the steps as described, wire bonding is performed in the first bonding point and the second bonding point. However, in the case of using the above-described wire bonding device of the related art, it is likely that the inertia of the bonding arm is increased due to extension of the bonding arm. Thus, the shape and the state of bonding become unstable because of impact load applied at the time of contact detection of the wire to the bonding subject, which may result in deteriorating the bondability. In other words, if the inertia of the bonding arm increases, the impact load applied at the time of the contact detection of the wire to the bonding subject increases, which leads to unstable shape and state of the bonding ball in the first bonding point and unstable crushed shape of the wire and the state in the second bonding point. Thus, there may be a case where the bondability is deteriorated. Furthermore, it is necessary to reduce the impact load by reducing the dropping speed at the time of contact detection in order to obtain a minute and highly precise bonding shape. Thereby, the productivity is to be deteriorated. An object of the present invention is to provide a wire bonding device and a wire bonding method in which the impact load applied at the time of contact detection of the wire to the bonding subject can be reduced and a stable and highly precise fabrication of a minute bonding ball can be achieved while enabling to shorten the bonding time. In order to achieve above mentioned object, a wire bonding device according to present invention for performing wire bonding on a bonding subject by bringing a wire, which is being lead out from a capillary provided in the end portion of an arm supported to be capable of swinging to be in contact with the bonding subject, wherein the arm comprises: a first arm supported by a first shaft to be capable of swinging; and a second arm with a smaller inertia than the first arm having the capillary in its end portion, which is supported to be capable of swinging by a second shaft provided on one end of the first arm. In the above-described bonding device of the present invention, a second arm having a capillary is supported to be capable of swinging on one end of the first arm supported to be capable of swinging. The inertia of the second arm is smaller than the inertia of the first arm. Therefore, by performing bonding through driving only the second arm at the time of bringing wire to be in contact with the bonding subject, the impact load applied at the time of the contact detection of the wire to the bonding subject can be reduced. As a result, compared to the case where the inertia of the arm is larger, the speed at the time of the contact detection of the wire to the bonding subject can be increased. Moreover, the wire bonding device according to present invention, further comprising a first driving unit for swinging the first arm and a second driving unit for swinging the second arm. A wire bonding method according to present invention for performing wire bonding by bringing a wire, which is being lead out from a capillary, to be in contact with a bonding subject by using a wire bonding device comprising an arm having: a first arm supported by a first shaft to be capable of swinging; and a second arm with a smaller inertia than the first arm having the capillary in its end portion, which is supported to be capable of swinging by a second shaft provided on one end of the first arm, the method comprising the steps of: bringing the capillary near the bonding subject by swinging the first arm in the approaching direction which is the direction of the capillary approaching the bonding subject until the wire reaches a prescribed height which is a position right before coming into a contact with the bonding subject; moving the capillary positioned in the height right before the contact by swinging the second arm in the approaching direction so that the wire is moved from the height right before the contact to be in contact with the bonding subject; and pressing the wire being in contact with the bonding subject against the bonding subject by swinging the second arm in the approaching direction in pressing step. In the above-described wire bonding method of the present invention, the capillary is brought near the bonding subject by swinging the first arm until reaching the height right before coming to be in contact with the bonding subject. Contacting of the wire to the bonding subject and pressurizing of the wire are performed by the second arm with the smaller inertia than that of the first arm. Thus, the impact load applied at the time of contact detection of the wire to the bonding subject can be reduced. Also, compared to the case where the inertia of the arm is larger, the speed at the time of the contact detection of the wire to the bonding subject can be increased. Moreover, the wire bonding method according to present invention, wherein the wire is bonded to the bonding subject by applying ultrasonic vibrations to the wire in the pressing step. Furthermore, the wire bonding method according to present invention, further comprising the steps of: after completing the bonding of the wire to the bonding subject, bringing the capillary back to the height right before the contact by swinging the second arm in the direction where the capillary recedes from the bonding subject. | Mid | [
0.5995975855130781,
37.25,
24.875
] |
The usefulness of EEG, exogenous evoked potentials, and cognitive evoked potentials in the acute stage of post-anoxic and post-traumatic coma. Three-modality evoked potentials (TMEPs) have been used for several years in association with the EEG as a diagnostic and prognostic tool in acute anoxic or traumatic coma. Cognitive EPs have been recently introduced. EEG and cognitive EPs provide functional assessment of the cerebral cortex. TMEP parameters can be described by two indices: the index of global cortical function (IGCF) and the index of brainstem conduction (IBSC). Although it remains a unique tool for epilepsy assessment, the value of EEG is largely limited by its high sensitivity to the electrical environmental noise, its dependence on sedative drugs, and its inability to test the brainstem. Major TMEP alterations (absence of cortical activities more than 24 hours after the onset of post-anoxic coma, major pontine involvement in head trauma) are associated in all cases with an ominous prognosis (death or vegetative state). However, even if mild TMEP changes are associated with a good prognosis in 65% (post-anoxic coma) to 90% (head trauma) of cases, some patients never recover despite exogenous TMEPs that are only mildly altered in the acute stage. Thus, cognitive EPs can usefully complement exogenous EPs as a prognostic tool in coma. Indeed, even if the absence of cognitive EPs in comatose patients does not have any prognostic value, their presence implies a very high (more than 90%) probability of consciousness recovery. The major technical challenge for the future will be the development of reliable tools for continuous EEG and TMEP monitoring. | High | [
0.673076923076923,
35,
17
] |
A SNEAKY thief stole a 50 inch television from Fearon Hall on Friday, August 15. The community centre, in Rectory Road, believes the theft happened some time after 9pm when everyone had left. Chantell Poxon from the centre told the Echo: “Someone was using upstairs and we think somebody’s hidden downstairs. There was an attempt about a month ago where somebody hid in the kitchen area and when everything was all locked up they tried to pull the TV off the wall but it’s chained to the wall so they got away with nothing, they just made a bit of a mess. It looks like they used something sharp to cut through the chain this time and left through the fire exit door. “Crowded House Church use it on a Sunday and the elderly people who come for community lunches use it as well. We probably won’t be able to afford to replace it. “These people don’t think about what they are doing, they just see an opportunity and go for it.” Loughborough Police commander Insp Nathan Kirk said there had been a recent “theme” of television thefts in the area, and urged anyone with information to come forward. “We’re not sure whether it’s one or two people responsible who are travelling or little pockets of criminals hitting us but we are doing work on it,” he said. “Anyone who has been offered second hand televisions, first of all be mindful that they might be stolen property but if you can pass any information about who is offering them we would be grateful for that.” Anyone who may have seen suspicious activity, or who has any information about a crime, is asked to contact police on 101 or Crimestoppers on 0800 555 111. | Low | [
0.519067796610169,
30.625,
28.375
] |
404 So.2d 1156 (1981) CITY OF CLEARWATER (FIRE DEPARTMENT), Appellant, v. John J. LEWIS, III, International Association of Fire Fighters, Local 1158, and the Public Employees Relations Commission, Appellees. No. 80-1715. District Court of Appeal of Florida, Second District. October 21, 1981. *1157 Frank Kowalski, Chief Asst. City Atty., Clearwater, for appellant. Rodney W. Smith, Gainesville, for appellees Lewis and International Association of Fire Fighters, Local 1158. Vernon Townes Grizzard and N. Sherrill Newton, Tallahassee, for appellee Public Employees Relations Commission. GRIMES, Judge. The City of Clearwater seeks review of a determination by the Public Employees Relations Commission (PERC) that it committed an unfair labor practice with respect to the termination of one of its employees. The city was charged with violating section 447.501(1)(a) and (b), Florida Statutes (1977), by refusing to permit John J. Lewis, III, to consult with a union representative at a meeting which resulted in his termination of employment. After the taking of testimony, a hearing officer made the following pertinent findings of fact. The city hired Lewis as a probationary fire fighter on February 14, 1977, and continued him as a probationary employee at all times during his employment. On February 15, 1978, Chief Pitts and Assistant Chief King called him into the chief's office. The chief told him that it was necessary to terminate his employment and gave as one of the reasons several of his performance evaluations. Chief Pitts gave him the option of resigning or being dismissed. The chief had decided to terminate Lewis before the meeting but called the meeting in order to review the performance evaluations with him and to give Lewis the option of resignation in order to protect his employment record. Lewis asked Chief Pitts if he could think over the choices and speak to the union. The chief told Lewis that the decision to terminate him had already been made and that consultation with a union representative would not change the decision. Lewis then signed a resignation form which the chief's secretary had prepared in advance. In his recommended order, the hearing officer observed that Chief Pitts could not deny Lewis the right to the presence of a union representative at the interview simply because Lewis occupied a probationary status. He also pointed out that Lewis had made a valid request for representation. *1158 However, the hearing officer concluded that the chief had not refused Lewis the right to have a union representative present, because he merely told Lewis that it wouldn't change his decision, and Lewis then decided to forego his right to representation and sign the resignation papers. PERC adopted the hearing officer's findings of fact with one exception not pertinent to this review.[1] However, by a split decision, PERC concluded that on these facts the city had committed an unfair labor practice by continuing the meeting after Lewis had requested union representation. PERC reasoned as follows: When Lewis was initially ordered to report downtown, and at the outset of the meeting with Chief Pitts, he had no indication of what was to transpire, and no reason to believe he might be disciplined. Thus, he had no reason to request union representation at that point. But when the Chief presented Lewis with the option of resigning or being dismissed, he posed a choice which Lewis could reasonably have believed might have significant adverse impact on his job interests and employment record. For example, whether Lewis resigned or was dismissed may well have been dispositive of his right to grieve under the collective bargaining agreement in effect between the parties. § 447.401, Fla. Stat. (1979); see generally Leon County CTA v. Leon County School Board, 6 FPER ¶ 11001 (1979). Second, Lewis' right to appeal under civil service was likewise implicated, a right which may also have depended on whether Lewis attained permanent employee status by the time of the meeting with Chief Pitts. § 447.401, Fla. Stat. (1979). Thus, the alternative of resignation, while protecting Lewis' employment record from the possible stigma of a dismissal, might have foreclosed the possibility of his remaining on the job by challenging successfully the grounds for his dismissal. These employment interests were at stake as Lewis contemplated the alternatives presented him by the Chief. To request union representation when making that choice, is we believe, "to engage in concerted activities not prohibited by law, for the purpose of... mutual aid or protection." § 447.301(3), Fla. Stat. (1979). As the United States Supreme Court stated in an analogous situation in NLRB v. Weingarten, Inc., 420 U.S. 251, 260-61, 95 S.Ct. 959, 965, 43 L.Ed.2d 171 (1975), this is true even though the employee alone may have an immediate stake in the outcome: The union representative whose participation he seeks is, however, safeguarding not only the particular employee's interest, but also the interests of the entire bargaining unit by exercising vigilance to make certain that the employer does not initiate or continue a practice of imposing punishment unjustly. The representative's presence is an assurance to other employees in the bargaining unit that they, too, can obtain his aid and protection if called upon to attend a like interview. (Footnote omitted.) All employees have an interest in knowing that they may request the presence of a union representative should they be confronted with the option of resigning or being dismissed. After being offered the option of resigning or being dismissed, Lewis asked to contact the union. The Hearing Officer did not find as a fact that Lewis ever withdrew his request for union representation. *1159 As the Hearing Officer analyzed the case, the fact that Lewis remained and signed resignation papers after the Chief told him the presence of a union representative would not change anything implies that Lewis decided to waive his request. The Commission disagrees with the Hearing Officer's analysis on this point. In another context, the Commission has held that waiver must be "clear and unmistakable." Palowitch v. Orange County School Board, 3 FPER 280, 282 (1977), aff'd, 367 So.2d 730 (Fla. 4th DCA 1979). With regard to this question, a commentator has written: Before inferring that a waiver has occurred ... the Board must assure itself that the employee acted knowingly and voluntarily. The right being waived is designed to prevent intimidation by the employer. It would be incongruous to infer a waiver without a clear indication that the very tactics the right is meant to prevent were not used to coerce a surrender of protection. Comment, "Union Presence in Disciplinary Meetings," 41 U.Chi.L.Rev. 329, 350 (1974) (cited by United States Supreme Court in NLRB v. J. Weingarten, Inc., supra). For the reasons discussed below, the Commission concludes that the factual circumstances of this case do not demonstrate such a "clear and unmistakable" waiver. The Commission has not disturbed the Hearing Officer's finding of fact that "Chief Pitts did not refuse the Charging Party permission to contact the union or have a union representative present" at the interview. Nevertheless, after Lewis once made a request for union representation, the Chief avoided a direct response, continued the meeting without responding to Lewis' request, and told Lewis that representation would not change anything. Under these circumstances, the Commission concludes that Lewis' failure to persist in his request for union representation does not constitute a "clear and unmistakable" waiver of representation. Accord, Southwestern Bell Telephone Co., 227 NLRB 1223 (1977). In Seitz v. Duval County School Board, 4 FPER ¶ 4154 (1978), rev'd in part on other grounds, 366 So.2d 119 (Fla. 1st DCA 1979), the Commission adopted the rationale of NLRB v. J. Weingarten, Inc., supra, thus guaranteeing to public employees the right to union representation in an investigatory or disciplinary interview if an employee has a reasonable belief that disciplinary action may result from the interview. The instant case is somewhat different from factual situations governed by the Weingarten case. Lewis' request for union representation was prompted not by apprehension that he might be disciplined, as was the case in Weingarten and Seitz, but by his concern that significant adverse consequences would result from the choice between resignation and dismissal. In both situations, however, the right to union representation arises from the Section 447.301(3) right to act for "mutual aid or protection." Once the right to representation attaches, and a request is made, the duty of the public employer is the same. Forthright guidelines for the future conduct of public employers, public employees, and employee organizations can best be provided by implementing this right to union representation in the following manner: When a public employer offers an employee an option to resign or be dismissed at a meeting with the employer, and the employee requests union representation, the employer is permitted one of three alternatives.[[2]] (1) grant the request; (2) discontinue the meeting; or (3) offer the employee the choice of continuing the meeting unaccompanied by a union representative or having no meeting at all. *1160 An employer may not continue the meeting in the absence of requested union representation unless the employee voluntarily and expressly agrees to remain unrepresented after the employer presents the employee with the choices mentioned in alternative (3) above. See United States Postal Service, 241 NLRB No. 18 (1979). An employer, under alternative (2) above, retains the right to revoke the option of resignation, discontinue the meeting, and dismiss the employee in accordance with his previous decision to terminate the employee's employment. The Commission is satisfied that neither the letter nor the spirit of this requirement was observed in the instant case. Chief Pitts did not grant Lewis' request for union representation, did not discontinue the meeting, and did not offer Lewis the choice of continuing the meeting unaccompanied by a union representative or having no meeting at all. Instead, Chief Pitts, having proposed the option of resignation, told Lewis that the presence of a union representative would not do him any good, since the decision to terminate Lewis' employment either by dismissal or by Lewis' resignation had already been made. In so conducting this meeting, Chief Pitts engaged in an unfair labor practice in violation of Section 447.501(1)(a), Florida Statutes (1979). (Footnotes omitted.) PERC revoked Lewis' resignation and ordered the city to present him with the option of resigning or being dismissed effective retroactively to February 15, 1978. The city raises two points which merit discussion. First, it contends that PERC erred in concluding that an employee has a right to union representation at a non-investigatory meeting at which he is simply given the option to resign or be dismissed. It also argues that PERC improperly overruled the hearing officer's finding that Lewis had waived his right to union representation when there was evidence in the record to support the findings. We will treat the points in the order stated. In NLRB v. J. Weingarten, Inc., 420 U.S. 251, 95 S.Ct. 959, 43 L.Ed.2d 171 (1975),[3] the court considered whether the denial of an employee's request to have a union representative present during an interview concerning the employee's suspected theft constituted an unfair labor practice. The Supreme Court reversed the court of appeals and reinstated a determination by the National Labor Relations Board that the action of the employee in seeking to have the assistance of a union representative at the conference with his employer fell within the wording of the National Labor Relations Act which provides that employees have the right to engage in concerted activities for the purpose of mutual aid and protection.[4] The Court upheld NLRB's construction that an employee has a right to have union representation at an interview where he reasonably believes that the investigation will result in disciplinary action. On the other hand, the NLRB has recently held that an employee has no right to union representation at a meeting with his employer held solely for the purpose of informing the employee of a previously determined disciplinary decision. Baton Rouge Water Works, 246 NLRB 161 (1979).[5] At the same time the Board stated: We stress that we are not holding today that there is no right to the presence of a union representative at any "disciplinary" interview. Indeed, if the employer engages in any conduct beyond merely informing the employee of a previously made disciplinary decision, the full panoply *1161 of protections accorded the employee under Weingarten may be applicable. Id. at 7. PERC's order goes beyond Weingarten because PERC adopted the hearing officer's finding that Chief Pitts had made up his mind to fire Lewis before the meeting. Therefore, the meeting was not an investigatory interview which might result in disciplinary action as contemplated by Weingarten. On the other hand, Chief Pitts' meeting with Lewis was something more than one to advise Lewis of a predetermined decision to fire him because the chief was also offering Lewis the option to resign. In the words of PERC, Lewis' request for union representation was prompted "not by apprehension that he might be disciplined" but rather "by his concern that significant adverse consequences would result from a choice between resignation and dismissal." PERC reasoned that in both situations the entitlement to union representation arose from the statutory "right to act for `mutual aid and protection.'" PERC Commissioner Parrish in his partially dissenting opinion in this case made the argument that Lewis was not entitled to the Weingarten type of protection when he said: Suffice it to say that on the facts in this case the Charging Party did not have a statutory right to the presence of a union representative at his termination meeting. The Charging Party was not forced to participate in an investigatory interview. Nor was he forced to disclose any information or take any action which was inimical to his job interests or to those of his fellow employees. The decision to terminate his employment had been made prior to the meeting and the meeting was held solely for the purposes of advising Lewis of the termination, advising him of the reasons for the termination, and advising him that he had an opportunity to resign in lieu of being terminated. Had the Charging Party been "forced" to choose between termination or resignation I might view this case otherwise. However, there is nothing in the findings of fact upon which to conclude that Respondent in any way coerced or intimated the Charging Party into signing a resignation document. Obviously Lewis was faced with a hard choice, but a hard choice is better than none at all. Further, it strains logic to conclude that an employer, by giving an employee an option to which the employee has no legal entitlement, thereby creates a situation in which the employer is vulnerable to a charge that it has deprived an employee of a right. It cannot be doubted that an option to resign in the face of certain termination can often be an option much sought after and desired by employees, and it is predictable that as a result of the majority's formulation of an interpretation of Section 447.301(3), Florida Statutes (1979), which exposes employers to Weingarten-type liability for offering their employees an opportunity to resign, such opportunities will become quite scarce in the future. In reaching our decision, this court must first determine whether to independently decide if the Florida Public Employees Labor Relations Act gives an employee the right to union representation at an interview in which he is given the option to be fired or resign or whether to accept PERC's determination that he has such a right. Reference to the act suggests that we should defer to the Commission. The statement of policy expressly charges PERC with "promoting harmonious and co-operative relationships between government and its employees, both collectively and individually." § 447.201, Fla. Stat. (1977). It was specifically created to "assist in resolving disputes between public employees and public employers." § 447.201(3), Fla. Stat. (1977). In remedying unfair labor practices, it is entitled to order such action "as will best implement the general policies expressed in this part." § 447.503(6)(a), Fla. Stat. (1977). These provisions indicate a legislative intent to delegate to PERC a range of discretion within *1162 which to make policy determinations necessarily involved in the interpretation and application of the "mutual aid and protection" guarantee set forth in section 447.301(3). Thus, an expert tribunal such as PERC is entitled to substantial deference in recognition of its special competence in dealing with labor problems. It is not our province to displace its choice between two conflicting views simply because we would have been justified in deciding the issue differently were it before us in the first instance. See Pasco County School Board v. PERC, 353 So.2d 108 (Fla. 1st DCA 1977). PERC was making policy when it determined that the Weingarten protections are applicable to situations in which an employee is given an option to resign or be fired. Commissioner Parrish persuasively suggests that the adoption of this policy may result in fewer employees being given the option to resign. On the other hand, an employee's decision to resign rather than be fired can affect his substantial rights,[6] and the advice of a union representative who is knowledgeable in these matters is likely to result in a more informed decision. Since there is a reasonable basis for the policy established by PERC in this case which is consistent with the philosophy of chapter 447, we will not disturb it. Cf. School Board v. PERC, 399 So.2d 520 (Fla. 2d DCA 1981), in which this court overturned a PERC order deemed to be arbitrary and out of harmony with the Public Employees Labor Relations Act. We also agree with PERC's contention that it did not improperly overrule the hearing officer's findings of fact on waiver. PERC simply held that the facts as found by the hearing officer did not constitute a waiver by Lewis of his right to union representation. Even Commissioner Parrish agreed that Lewis' action of continuing with the meeting and ultimately signing the resignation did not constitute a clear and unmistakable waiver of his right to union representation. To the extent that the determination of whether a waiver occurred could be considered to be a question of fact, it is the kind of factual issue on which an agency has more discretion to overrule a hearing examiner in the implementation of agency policy. Thus, in McDonald v. Department of Banking & Finance, 346 So.2d 569 (Fla. 1st DCA 1977), the court wrote: In determining whether substantial evidence supports the agency's substituted findings of fact, a reviewing court will naturally accord greater probative force to the hearing officer's contrary findings when the question is simply the weight or credibility of testimony by witnesses, or when the factual issues are otherwise susceptible of ordinary methods of proof, or when concerning those facts the agency may not rightfully claim special insight. At the other end of the scale, where the ultimate facts are increasingly matters of opinion and opinions are increasingly infused by policy considerations for which the agency has special responsibility, a reviewing court will give correspondingly less weight to the hearing officer's findings in determining the substantiality of evidence supporting the agency's substituted findings. Thus, the substantiality of evidence supporting an agency's substituted finding of fact depends on a number of variables: how susceptible is the factual issue to resolution by credible witnesses and other evidence, how substantially the hearing officer's discarded findings are supported by such evidence, how far the factual issue tends to be one of opinion, how completely agency policy occupies a field otherwise open to different opinion. Id. at 579.[7] Given the fact that an employee under these circumstances has the right to union representation once he makes the request, PERC also had the authority to *1163 establish a standard for the waiver of this right which is somewhat akin to that required for the waiver of counsel in criminal cases.[8] Accordingly, we deny the petition for review. SCHEB, C.J., and DANAHY, J., concur. NOTES [1] Clearwater's probationary period for firemen is one year. Since the interview took place one day after Lewis had worked for a year, there was some question concerning his probationary status. Therefore, PERC modified the hearing officer's findings to eliminate the determination that Lewis was still on probationary status. As modified, the findings reflected that Lewis was employed by the city and that as of his termination the city had taken no formal personnel action to change his designation from probationary to permanent. The city quarrels about this but cites us no authority for the proposition that Lewis' right to the advice of a union representative (as contrasted to his right of tenure) was different depending on whether he occupied a probationary or permanent employment status. See § 447.203(3), Fla. Stat. (1977). [2] In a footnote PERC pointed out that the National Labor Relations Board had first formulated these alternatives in the context of investigatory or disciplinary interviews citing Roadway Express, Inc., 246 NLRB 180 (1979); United States Postal Serv., 241 NLRB 18 (1979); General Elec. Co., 240 NLRB 66 (1979). [3] In interpreting Florida's Public Employees Labor Relations Act, decisions construing similar provisions of the National Labor Relations Act are persuasive but not binding. Pasco County School Bd. v. PERC, 353 So.2d 108, 116 (Fla. 1st DCA 1977). [4] The Florida Public Employees Labor Relations Act contains similar language. § 447.301(3), Fla. Stat. (1977). [5] The Fifth Circuit Court of Appeals has approved the rationale of Baton Rouge Water Works in Anchortank, Inc. v. NLRB, 618 F.2d 1153 (5th Cir.1980). [6] In addition to those outlined in PERC's order, Lewis' decision to resign may have had a very real impact upon his right to claim unemployment compensation. See § 443.06, Fla. Stat. (1977). [7] We adopted these principles in Koltay v. Division of Gen. Reg., 374 So.2d 1386 (Fla. 2d DCA 1979). [8] However, unlike the criminal procedures, we know of no authority which would require the employer to advise the employee of his right to union representation in the first place. | Mid | [
0.5460385438972161,
31.875,
26.5
] |
/* * linux/drivers/ide/pci/sl82c105.c * * SL82C105/Winbond 553 IDE driver * * Maintainer unknown. * * Drive tuning added from Rebel.com's kernel sources * -- Russell King (15/11/98) [email protected] * * Merge in Russell's HW workarounds, fix various problems * with the timing registers setup. * -- Benjamin Herrenschmidt (01/11/03) [email protected] * * Copyright (C) 2006-2007 MontaVista Software, Inc. <[email protected]> */ #include <linux/types.h> #include <linux/module.h> #include <linux/kernel.h> #include <linux/timer.h> #include <linux/mm.h> #include <linux/ioport.h> #include <linux/interrupt.h> #include <linux/blkdev.h> #include <linux/hdreg.h> #include <linux/pci.h> #include <linux/ide.h> #include <asm/io.h> #include <asm/dma.h> #undef DEBUG #ifdef DEBUG #define DBG(arg) printk arg #else #define DBG(fmt,...) #endif /* * SL82C105 PCI config register 0x40 bits. */ #define CTRL_IDE_IRQB (1 << 30) #define CTRL_IDE_IRQA (1 << 28) #define CTRL_LEGIRQ (1 << 11) #define CTRL_P1F16 (1 << 5) #define CTRL_P1EN (1 << 4) #define CTRL_P0F16 (1 << 1) #define CTRL_P0EN (1 << 0) /* * Convert a PIO mode and cycle time to the required on/off times * for the interface. This has protection against runaway timings. */ static unsigned int get_pio_timings(ide_pio_data_t *p) { unsigned int cmd_on, cmd_off; cmd_on = (ide_pio_timings[p->pio_mode].active_time + 29) / 30; cmd_off = (p->cycle_time - 30 * cmd_on + 29) / 30; if (cmd_on == 0) cmd_on = 1; if (cmd_off == 0) cmd_off = 1; return (cmd_on - 1) << 8 | (cmd_off - 1) | (p->use_iordy ? 0x40 : 0x00); } /* * Configure the chipset for PIO mode. */ static u8 sl82c105_tune_pio(ide_drive_t *drive, u8 pio) { struct pci_dev *dev = HWIF(drive)->pci_dev; int reg = 0x44 + drive->dn * 4; ide_pio_data_t p; u16 drv_ctrl; DBG(("sl82c105_tune_pio(drive:%s, pio:%u)\n", drive->name, pio)); pio = ide_get_best_pio_mode(drive, pio, 5, &p); drv_ctrl = get_pio_timings(&p); /* * Store the PIO timings so that we can restore them * in case DMA will be turned off... */ drive->drive_data &= 0xffff0000; drive->drive_data |= drv_ctrl; if (!drive->using_dma) { /* * If we are actually using MW DMA, then we can not * reprogram the interface drive control register. */ pci_write_config_word(dev, reg, drv_ctrl); pci_read_config_word (dev, reg, &drv_ctrl); } printk(KERN_DEBUG "%s: selected %s (%dns) (%04X)\n", drive->name, ide_xfer_verbose(pio + XFER_PIO_0), p.cycle_time, drv_ctrl); return pio; } /* * Configure the drive and chipset for a new transfer speed. */ static int sl82c105_tune_chipset(ide_drive_t *drive, u8 speed) { static u16 mwdma_timings[] = {0x0707, 0x0201, 0x0200}; u16 drv_ctrl; DBG(("sl82c105_tune_chipset(drive:%s, speed:%s)\n", drive->name, ide_xfer_verbose(speed))); speed = ide_rate_filter(drive, speed); switch (speed) { case XFER_MW_DMA_2: case XFER_MW_DMA_1: case XFER_MW_DMA_0: drv_ctrl = mwdma_timings[speed - XFER_MW_DMA_0]; /* * Store the DMA timings so that we can actually program * them when DMA will be turned on... */ drive->drive_data &= 0x0000ffff; drive->drive_data |= (unsigned long)drv_ctrl << 16; /* * If we are already using DMA, we just reprogram * the drive control register. */ if (drive->using_dma) { struct pci_dev *dev = HWIF(drive)->pci_dev; int reg = 0x44 + drive->dn * 4; pci_write_config_word(dev, reg, drv_ctrl); } break; case XFER_PIO_5: case XFER_PIO_4: case XFER_PIO_3: case XFER_PIO_2: case XFER_PIO_1: case XFER_PIO_0: (void) sl82c105_tune_pio(drive, speed - XFER_PIO_0); break; default: return -1; } return ide_config_drive_speed(drive, speed); } /* * Check to see if the drive and chipset are capable of DMA mode. */ static int sl82c105_ide_dma_check(ide_drive_t *drive) { DBG(("sl82c105_ide_dma_check(drive:%s)\n", drive->name)); if (ide_tune_dma(drive)) return 0; return -1; } /* * The SL82C105 holds off all IDE interrupts while in DMA mode until * all DMA activity is completed. Sometimes this causes problems (eg, * when the drive wants to report an error condition). * * 0x7e is a "chip testing" register. Bit 2 resets the DMA controller * state machine. We need to kick this to work around various bugs. */ static inline void sl82c105_reset_host(struct pci_dev *dev) { u16 val; pci_read_config_word(dev, 0x7e, &val); pci_write_config_word(dev, 0x7e, val | (1 << 2)); pci_write_config_word(dev, 0x7e, val & ~(1 << 2)); } /* * If we get an IRQ timeout, it might be that the DMA state machine * got confused. Fix from Todd Inglett. Details from Winbond. * * This function is called when the IDE timer expires, the drive * indicates that it is READY, and we were waiting for DMA to complete. */ static int sl82c105_ide_dma_lostirq(ide_drive_t *drive) { ide_hwif_t *hwif = HWIF(drive); struct pci_dev *dev = hwif->pci_dev; u32 val, mask = hwif->channel ? CTRL_IDE_IRQB : CTRL_IDE_IRQA; u8 dma_cmd; printk("sl82c105: lost IRQ, resetting host\n"); /* * Check the raw interrupt from the drive. */ pci_read_config_dword(dev, 0x40, &val); if (val & mask) printk("sl82c105: drive was requesting IRQ, but host lost it\n"); /* * Was DMA enabled? If so, disable it - we're resetting the * host. The IDE layer will be handling the drive for us. */ dma_cmd = inb(hwif->dma_command); if (dma_cmd & 1) { outb(dma_cmd & ~1, hwif->dma_command); printk("sl82c105: DMA was enabled\n"); } sl82c105_reset_host(dev); /* __ide_dma_lostirq would return 1, so we do as well */ return 1; } /* * ATAPI devices can cause the SL82C105 DMA state machine to go gaga. * Winbond recommend that the DMA state machine is reset prior to * setting the bus master DMA enable bit. * * The generic IDE core will have disabled the BMEN bit before this * function is called. */ static void sl82c105_dma_start(ide_drive_t *drive) { ide_hwif_t *hwif = HWIF(drive); struct pci_dev *dev = hwif->pci_dev; sl82c105_reset_host(dev); ide_dma_start(drive); } static int sl82c105_ide_dma_timeout(ide_drive_t *drive) { ide_hwif_t *hwif = HWIF(drive); struct pci_dev *dev = hwif->pci_dev; DBG(("sl82c105_ide_dma_timeout(drive:%s)\n", drive->name)); sl82c105_reset_host(dev); return __ide_dma_timeout(drive); } static int sl82c105_ide_dma_on(ide_drive_t *drive) { struct pci_dev *dev = HWIF(drive)->pci_dev; int rc, reg = 0x44 + drive->dn * 4; DBG(("sl82c105_ide_dma_on(drive:%s)\n", drive->name)); rc = __ide_dma_on(drive); if (rc == 0) { pci_write_config_word(dev, reg, drive->drive_data >> 16); printk(KERN_INFO "%s: DMA enabled\n", drive->name); } return rc; } static void sl82c105_dma_off_quietly(ide_drive_t *drive) { struct pci_dev *dev = HWIF(drive)->pci_dev; int reg = 0x44 + drive->dn * 4; DBG(("sl82c105_dma_off_quietly(drive:%s)\n", drive->name)); pci_write_config_word(dev, reg, drive->drive_data); ide_dma_off_quietly(drive); } /* * Ok, that is nasty, but we must make sure the DMA timings * won't be used for a PIO access. The solution here is * to make sure the 16 bits mode is diabled on the channel * when DMA is enabled, thus causing the chip to use PIO0 * timings for those operations. */ static void sl82c105_selectproc(ide_drive_t *drive) { ide_hwif_t *hwif = HWIF(drive); struct pci_dev *dev = hwif->pci_dev; u32 val, old, mask; //DBG(("sl82c105_selectproc(drive:%s)\n", drive->name)); mask = hwif->channel ? CTRL_P1F16 : CTRL_P0F16; old = val = (u32)pci_get_drvdata(dev); if (drive->using_dma) val &= ~mask; else val |= mask; if (old != val) { pci_write_config_dword(dev, 0x40, val); pci_set_drvdata(dev, (void *)val); } } /* * ATA reset will clear the 16 bits mode in the control * register, we need to update our cache */ static void sl82c105_resetproc(ide_drive_t *drive) { struct pci_dev *dev = HWIF(drive)->pci_dev; u32 val; DBG(("sl82c105_resetproc(drive:%s)\n", drive->name)); pci_read_config_dword(dev, 0x40, &val); pci_set_drvdata(dev, (void *)val); } /* * We only deal with PIO mode here - DMA mode 'using_dma' is not * initialised at the point that this function is called. */ static void sl82c105_tune_drive(ide_drive_t *drive, u8 pio) { DBG(("sl82c105_tune_drive(drive:%s, pio:%u)\n", drive->name, pio)); pio = sl82c105_tune_pio(drive, pio); (void) ide_config_drive_speed(drive, XFER_PIO_0 + pio); } /* * Return the revision of the Winbond bridge * which this function is part of. */ static unsigned int sl82c105_bridge_revision(struct pci_dev *dev) { struct pci_dev *bridge; u8 rev; /* * The bridge should be part of the same device, but function 0. */ bridge = pci_get_bus_and_slot(dev->bus->number, PCI_DEVFN(PCI_SLOT(dev->devfn), 0)); if (!bridge) return -1; /* * Make sure it is a Winbond 553 and is an ISA bridge. */ if (bridge->vendor != PCI_VENDOR_ID_WINBOND || bridge->device != PCI_DEVICE_ID_WINBOND_83C553 || bridge->class >> 8 != PCI_CLASS_BRIDGE_ISA) { pci_dev_put(bridge); return -1; } /* * We need to find function 0's revision, not function 1 */ pci_read_config_byte(bridge, PCI_REVISION_ID, &rev); pci_dev_put(bridge); return rev; } /* * Enable the PCI device * * --BenH: It's arch fixup code that should enable channels that * have not been enabled by firmware. I decided we can still enable * channel 0 here at least, but channel 1 has to be enabled by * firmware or arch code. We still set both to 16 bits mode. */ static unsigned int __devinit init_chipset_sl82c105(struct pci_dev *dev, const char *msg) { u32 val; DBG(("init_chipset_sl82c105()\n")); pci_read_config_dword(dev, 0x40, &val); val |= CTRL_P0EN | CTRL_P0F16 | CTRL_P1F16; pci_write_config_dword(dev, 0x40, val); pci_set_drvdata(dev, (void *)val); return dev->irq; } /* * Initialise IDE channel */ static void __devinit init_hwif_sl82c105(ide_hwif_t *hwif) { unsigned int rev; DBG(("init_hwif_sl82c105(hwif: ide%d)\n", hwif->index)); hwif->tuneproc = &sl82c105_tune_drive; hwif->speedproc = &sl82c105_tune_chipset; hwif->selectproc = &sl82c105_selectproc; hwif->resetproc = &sl82c105_resetproc; /* * We support 32-bit I/O on this interface, and * it doesn't have problems with interrupts. */ hwif->drives[0].io_32bit = hwif->drives[1].io_32bit = 1; hwif->drives[0].unmask = hwif->drives[1].unmask = 1; /* * We always autotune PIO, this is done before DMA is checked, * so there's no risk of accidentally disabling DMA */ hwif->drives[0].autotune = hwif->drives[1].autotune = 1; if (!hwif->dma_base) return; rev = sl82c105_bridge_revision(hwif->pci_dev); if (rev <= 5) { /* * Never ever EVER under any circumstances enable * DMA when the bridge is this old. */ printk(" %s: Winbond W83C553 bridge revision %d, " "BM-DMA disabled\n", hwif->name, rev); return; } hwif->atapi_dma = 1; hwif->mwdma_mask = 0x07; hwif->ide_dma_check = &sl82c105_ide_dma_check; hwif->ide_dma_on = &sl82c105_ide_dma_on; hwif->dma_off_quietly = &sl82c105_dma_off_quietly; hwif->ide_dma_lostirq = &sl82c105_ide_dma_lostirq; hwif->dma_start = &sl82c105_dma_start; hwif->ide_dma_timeout = &sl82c105_ide_dma_timeout; if (!noautodma) hwif->autodma = 1; hwif->drives[0].autodma = hwif->drives[1].autodma = hwif->autodma; if (hwif->mate) hwif->serialized = hwif->mate->serialized = 1; } static ide_pci_device_t sl82c105_chipset __devinitdata = { .name = "W82C105", .init_chipset = init_chipset_sl82c105, .init_hwif = init_hwif_sl82c105, .channels = 2, .autodma = NOAUTODMA, .enablebits = {{0x40,0x01,0x01}, {0x40,0x10,0x10}}, .bootable = ON_BOARD, }; static int __devinit sl82c105_init_one(struct pci_dev *dev, const struct pci_device_id *id) { return ide_setup_pci_device(dev, &sl82c105_chipset); } static struct pci_device_id sl82c105_pci_tbl[] = { { PCI_DEVICE(PCI_VENDOR_ID_WINBOND, PCI_DEVICE_ID_WINBOND_82C105), 0}, { 0, }, }; MODULE_DEVICE_TABLE(pci, sl82c105_pci_tbl); static struct pci_driver driver = { .name = "W82C105_IDE", .id_table = sl82c105_pci_tbl, .probe = sl82c105_init_one, }; static int __init sl82c105_ide_init(void) { return ide_pci_register_driver(&driver); } module_init(sl82c105_ide_init); MODULE_DESCRIPTION("PCI driver module for W82C105 IDE"); MODULE_LICENSE("GPL"); | Low | [
0.48616600790513803,
30.75,
32.5
] |
Archived News OpenTTD 1.0.0 2010-04-01 Posted by frosch 6 years. What were you doing 6 years ago? In March 2004 OpenTTD 0.1 was released. Hardly a month later in April 2004 OpenTTD 0.2. And today, six years later... OpenTTD 1.0.0. It was a lot of work, hundreds of thousands of translations, tens of thousands of commits, thousands of graphics, hundreds of patches, dozen of sounds and musics, and one goal. How many people contributed? Dozen of artists, translators and developers, hundreds of testers and bug reporters, and also the thousands of players. Looking at the readmes and credit sections only gives a small hint. Some of those who were main contributors left long ago, and there are only a few who know them all and talked to them once via IRC or the forums. But if you consider all contributors, including those of the used libraries, and the external artists of OpenSFX... Well, then most likely not every contributor actually knows OpenTTD :) So in the end, what was most fun in the past 6 years of OpenTTD? Playing? Contributing? Modding? Talking? Or just taking part in a large crowed moving in one direction? One direction? Well, at least in bigger scope :) But in more detail there were quite some parties involved in the process. Sometimes pulling in the same direction, sometimes maybe pulling in slightly different ones. Let's just mention some of the projects around OpenTTD which influenced it in this or that direction: The various integrated builds and patchpacks, first of all the MiniIN. Then the first Town Growth Challenge, TTDPatch, #openttdcoop, Goal Servers and the big patches (Subsidiaries, YAPF, YAPP, CargoDist, 32bpp & ExtraZoom). And not everything which made it into main trunk was happy sunshine, just to mention the first approach to Path Based Signalling, or the attempts around the AI. But when looking back, most turned out fun. Thank you! | High | [
0.673728813559322,
39.75,
19.25
] |
Transversely pumped 11-pass amplifier for femtosecond optical pulses. An off-axis near-concentric interferometer is used for multiple passages through a small (2-mm-diameter) cylindrical volume of amplifying dye. The gain volume is pumped transversely by a copper-vapor laser, in a geometry designed for uniform pump. Pulse-to-pulse reproducibility better than 1% is obtained by operating with full saturation and using a solvent with a vanishing temperature-induced change in index of refraction (dn/dT = 0). | High | [
0.698140200286123,
30.5,
13.1875
] |
During an interview with conservative pundit Rush Limbaugh, US President George Bush expressed deep concerns about the possibility of the United States leaving the Middle East, raising fears that extremists could topple governments to "control oil resources." "Give me a second here, Rush, because I want to share something with you," Bush said. "I am deeply concerned about a country, the United States, leaving the Middle East." Bush said that he was "worried that rival forms of extremists will battle for power, obviously creating incredible damage if they do so; that they will topple modern governments, that they will be in a position to use oil as a tool to blackmail the West." "People say, 'What do you mean by that?'" The president continued. "I say, 'If they control oil resources, then they pull oil off the market in order to run the price up, and they will do so unless we abandon Israel, for example, or unless we abandon allies.'" "You couple that with a country that doesn't like us with a nuclear weapon, and people will look back at this moment and say, 'What happened to those people in 2006?' and those are the stakes in this war we face," Bush said. "On the one hand we've got a plan to make sure we protect you from immediate attack, and on the other hand we've got a long-term strategy to deal with these threats, and part of that strategy is to stay on the offense," Bush continued. "Part of the strategy is to help young democracies like Lebanon and Iraq be able to survive against the terrorists and the extremists who are trying to crush their hopes, and part of the democracy is for a freedom movement, which will help create the conditions so that the extremists become marginalized and unable to recruit." | Low | [
0.5263157894736841,
31.25,
28.125
] |
Q: pl/sql bubble sort okay, I am beating myself up over this. I am need to load a array in people last name stored in a table. Then sort the last names and print them out in alphabetical order. This must be done using the bubble sort algorithm. here is what I have so far CREATE OR REPLACE PROCEDURE TEAM_TABLE_SORT AS TYPE player_Name_type IS TABLE OF databasename.team.player%type INDEX BY PLS_INTEGER ; player_name player_Name_type; i integer := 1; temp integer; BEGIN FOR player_names IN (SELECT * FROM marshall.team ) LOOP player_name(i) := player_names.player; DBMS_OUTPUT.PUT_LINE(i|| ' - ' ||chr(9) || player_name(i) ) ; i := i + 1 ; END LOOP All this really does is print out the names. I cannot get it to sort. I am not try thing this TYPE player_Name_type IS TABLE OF %type INDEX BY varchar2(20) ; aux player_Name_type; i integer := 1; v_current is table of aux swapped BOOLEAN := TRUE; BEGIN FOR aux IN (SELECT * FROM ) LOOP DBMS_OUTPUT.PUT_LINE(i|| ' - ' ||chr(9) || aux.player); i := i + 1 ; END LOOP; v_current := aux.first; WHILE(swapped) LOOP swapped := FALSE; FOR I IN 1..(aux.count-2) LOOP IF aux(i) > aux(I+1) THEN v_current := aux(i+1); aux(I+1) := aux(i); aux(i) := v_current; END IF; swapped := TRUE; END LOOP; END LOOP; FOR aux IN (SELECT * FROM LOOP DBMS_OUTPUT.PUT_LINE(i|| ' - ' ||chr(9) ||aux.player); i := i + 1 ; END LOOP; A: This should be what you are looking for. Note that it is better to type the variables/collections off of the tables like you have in your example. I just used generic versions since I don't have your tables to work with. If you don't understand how this is working, feel free to ask. I am guessing this is homework (who else would bubble sort in Oracle), so the point of the assignment is for you to understand it, not just to get it right. :) DECLARE coll DBMS_SQL.VARCHAR2A; swapped BOOLEAN; tmp VARCHAR2(10); BEGIN /* Generate 10 random strings and collect them into our collection Note: you would replace this with your query on marshall.team */ select dbms_random.string('l',10) rand_string BULK COLLECT INTO coll from dual connect by level <= 10; /* At this point, all of the rows we need are in our collection so there is no need to go back to the table anymore. Now onto the... Bubble sort.. walk through the collection swapping elements until we make a pass where no swapping takes place */ LOOP swapped := false; FOR i IN 2 .. coll.LAST LOOP IF coll(i-1) > coll(i) THEN -- swap records tmp := coll(i); coll(i) := coll(i-1); coll(i-1) := tmp; /* Mark that swap has taken place. note we mark as true only inside the if block, meaning a swap really did take place */ swapped := true; END IF; END LOOP; -- If we passed through table without swapping we are done, so exit EXIT WHEN NOT swapped; END LOOP; /* Now print out records to make sure they are in order. Again notice how we are just referencing the (now sorted) collection and not going back to the table again */ FOR i in coll.FIRST .. coll.LAST LOOP dbms_output.put_line(coll(i)); END LOOP; END; / | Mid | [
0.5990990990990991,
33.25,
22.25
] |
Complications in the surgical treatment of carcinoma of the endometrium. The authors report their experience in medical and surgical complications after surgical treatment of endometrial carcinoma, from January 1976 to December 1992, 301 cases of adenocarcinoma were operated by abdominal or vaginal route. From 1980 onwards abdominal route was the most frequent (radical hysterectomy with bilateral adnexectomy Rutledge type II-III with pelvic and/or aortic lymphadenectomy). No lesion occurred either during surgery or later, in the urinary or intestinal apparatus or to the great abdomino-pelvic vessels. The only medical complication observed was one episode of cerebral ictus three days after operation. Two cases of adynamic ileus and five of ventral hernia occurred. | High | [
0.665796344647519,
31.875,
16
] |
Pharmacological profile of 48740 R.P., a PAF-acether antagonist. The pyrrolo-thiazole derivative 48740 R.P. inhibited the platelet-activating factor (PAF-acether)-induced aggregation of human and rabbit platelets and was poorly effective against ADP- and arachidonic acid-induced platelet aggregation. 48740 R.P. prevented the activation of guinea-pig alveolar macrophages by PAF-acether, and the PAF-acether-induced thromboxane B2 production from guinea-pig lungs. 48740 R.P. (3 mg/kg i.v.) antagonized selectively in anaesthetized guinea-pigs the bronchoconstriction due to PAF-acether without affecting that due to acetylcholine, histamine, serotonin, thromboxane A2 analogue U-46,619 and arachidonic acid. A higher dose of 48740 R.P. (10 mg/kg i.v.) was required to block the thrombocytopenia and the leucopenia induced by PAF-acether in the propranolol-treated guinea-pigs. 48740 R.P. (30 mg/kg i.v.) antagonized the PAF-acether effects when bronchoconstriction was induced by aerosolized PAF-acether. 48740 R.P. is a selective antagonist of PAF-acether under in vitro and in vivo conditions. | High | [
0.662125340599455,
30.375,
15.5
] |
MA Gutter Cleaning Company Cited for Repeat Safety Violations The Department of Labor’s Occupational Safety and Health Administration (OSHA) has levied fines against New Jersey-based Ned Stevens Gutter Cleaning and General Contracting of Massachusetts Inc. after a second incident involving a worker falling from a roof happened within the same year. The first incident occurred on Nov. 29, 2015 when an employee fell 26 feet off a roof in Newton, and the second incident occurred in Lexington on Oct. 24, 2016 when a worker fell nine feet from a garage roof. The same company had already been cited for a similar event in March of 2014 as well that occurred in New Jersey. An inspection by OSHA revealed that, in the incident that occurred in Lexington, the employee working atop the garage did not have adequate fall protection prior to his accident, and neither did his foreman, who was vulnerable to a fall from a much higher roof of 26 feet. OSHA fined Ned Stevens $68,591 in total for this violation. As a result of the incident in Newton that occurred less than a year prior, OSHA levied $45,500 in fines for failing to ensure that employees had the proper safety equipment to safely perform their work duties. In this incident, once again, workers were working on tall rooftops without fall protection. Despite the conclusiveness of the investigation, Ned Stevens is contesting these penalties. “This is a disturbing pattern – two falls, two incidents – in which this employer did not provide its employees with the proper and required safeguards. The seriousness of fall hazards cannot be understated,” said Anthony Covello, OSHA’s area director for Middlesex and Essex counties. “The Bureau of Labor Statistics reports 10 workers in Massachusetts died in falls in 2014. While the workers in these two incidents were not killed, the threat of death or disabling injuries was real and present. Ned Stevens must act to protect its employees against this common but preventable hazard at all its jobsites.” Safety protection is not an option If you are an employer that deals specifically with business that involves cleaning, installing or repairing gutters – which obviously will involve employees being exposed to the risk of falls – then providing safety precautions to prevent a serious injury (or death) while on the job is not just simply the right thing to do morally, it is a legal requirement under the Occupational Safety and Health Act of 1970. Employers who fail to provide proper safety precautions to workers who work in hazardous conditions – such as high up on rooftops – will face serious financial penalties from OSHA, and can additionally face much steeper penalties in lawsuits from employees (or their family members) that are injured while on the job. Falls from high places are consistently one of the most common causes of work-related fatalities in the United States, and the stakes are too high to not properly equip workers who must perform duties under these conditions. Let us fight on your behalf If you or a loved one was injured or killed due to a serious fall at work because they were not given adequate safety protection – such as fall prevention gear – then you are entitled to file a claim against the employer who failed to provide a safe work environment. Our team of legal professionals at Altman & Altman LLP have extensive experience litigating against negligent employers who cause harm to their employees, and we will fight for you to hold them responsible and secure the financial compensation you need to pay your hospital bills, pay for recovery costs and restore wages that were lost from being unable to work. Call us for a free consultation today at 617-492-3000 or toll-free at 800-481-6199. We are available 24/7 and we don’t collect a penny unless you are successful in your claim. Some Select Cases are referred to other attorneys for principal responsibility By publishing this information on this Website, the Boston, Massachusetts law firm of Altman & Altman LLP is not claiming to represent any clients or cases mentioned here. The content provided is designed to inform readers and is not intended as legal advice. | Mid | [
0.596330275229357,
32.5,
22
] |
/** * \file * \brief Defines an execution handler that executes a given command on a * device and returns the results. * * The basic flow is to wake the device, send the command, wait/poll for * completion, and finally receives the response from the device and does * basic checks before returning to caller. * * This handler supports the ATSHA and ATECC device family. * * \copyright (c) 2015-2020 Microchip Technology Inc. and its subsidiaries. * * \page License * * Subject to your compliance with these terms, you may use Microchip software * and any derivatives exclusively with Microchip products. It is your * responsibility to comply with third party license terms applicable to your * use of third party software (including open source software) that may * accompany Microchip software. * * THIS SOFTWARE IS SUPPLIED BY MICROCHIP "AS IS". NO WARRANTIES, WHETHER * EXPRESS, IMPLIED OR STATUTORY, APPLY TO THIS SOFTWARE, INCLUDING ANY IMPLIED * WARRANTIES OF NON-INFRINGEMENT, MERCHANTABILITY, AND FITNESS FOR A * PARTICULAR PURPOSE. IN NO EVENT WILL MICROCHIP BE LIABLE FOR ANY INDIRECT, * SPECIAL, PUNITIVE, INCIDENTAL OR CONSEQUENTIAL LOSS, DAMAGE, COST OR EXPENSE * OF ANY KIND WHATSOEVER RELATED TO THE SOFTWARE, HOWEVER CAUSED, EVEN IF * MICROCHIP HAS BEEN ADVISED OF THE POSSIBILITY OR THE DAMAGES ARE * FORESEEABLE. TO THE FULLEST EXTENT ALLOWED BY LAW, MICROCHIP'S TOTAL * LIABILITY ON ALL CLAIMS IN ANY WAY RELATED TO THIS SOFTWARE WILL NOT EXCEED * THE AMOUNT OF FEES, IF ANY, THAT YOU HAVE PAID DIRECTLY TO MICROCHIP FOR * THIS SOFTWARE. */ #ifndef CALIB_EXECUTION_H #define CALIB_EXECUTION_H #include "atca_status.h" #include "calib_command.h" #include "atca_device.h" #include "atca_config.h" #ifdef __cplusplus extern "C" { #endif #define ATCA_UNSUPPORTED_CMD ((uint16_t)0xFFFF) #ifdef ATCA_NO_POLL /** \brief Structure to hold the device execution time and the opcode for the * corresponding command */ typedef struct { uint8_t opcode; uint16_t execution_time_msec; }device_execution_time_t; ATCA_STATUS calib_get_execution_time(uint8_t opcode, ATCACommand ca_cmd); #endif ATCA_STATUS calib_execute_command(ATCAPacket* packet, ATCADevice device); #ifdef __cplusplus } #endif #endif | Mid | [
0.6350515463917521,
38.5,
22.125
] |
Effect of salt stress on morpho-physiology, vegetative growth and yield of rice. Selection of salt tolerant rice varieties has a huge impact on global food supply chain. Five Malaysian rice (Oryza sativa L.) varieties, MR33, MR52, MR211, MR219 and MR232 were tested in pot experiment under different salinity levels for their response in term of vegetative growth, physiological activities, development of yield components and grain yield. Rice varieties, BRRI dhan29 and IR20 were used as a salt-sensitive control and Pokkali was used as a salt-tolerant control. Three different salinity levels viz. 4, 8, and 12 dS m(-1) were used in a randomized complete block design with four replications under glass house conditions. Two Malaysia varieties, MR211 and MR232 performed better in terms of vegetative growth (plant height, leaf area plant(-1), number of tillers plant(-1), dry matter accumulation plant(-1)), photosynthetic rate, transpiration rate, yield components, grain yield and injury symptoms. While, MR33, MR52 and MR219 verities were able to withstand salinity stress over salt-sensitive control, BRRI dhan29 and IR20. | Mid | [
0.6198979591836731,
30.375,
18.625
] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.