id
int64 393k
2.82B
| repo
stringclasses 68
values | title
stringlengths 1
936
| body
stringlengths 0
256k
⌀ | labels
stringlengths 2
508
| priority
stringclasses 3
values | severity
stringclasses 3
values |
|---|---|---|---|---|---|---|
284,767,792 | opencv | setWindowProperty will not put window into fullscreen mode unless window property set to WINDOW_NORMAL | ##### System information (version)
- OpenCV => 3.1
- Operating System / Platform => Linux Ubuntu 16.04
- Compiler => gcc
##### Detailed description
On GTK, calling `cv::setWindowProperty(name, WND_PROP_FULLSCREEN, WINDOW_FULLSCREEN)` will only enforce the window to go fullscreen *if* the current property value on that window is `WINDOW_NORMAL`). This is can also be observed in the highgui GTK source code [here](https://github.com/opencv/opencv/blob/9e6ef99422d5594dff5f30e4b6f1c695e78ce67e/modules/highgui/src/window_gtk.cpp#L744).
I discovered this after trying to force a window to fullscreen with `setWindowProperty`, since `cv::namedWindow(name, WINDOW_FULLSCREEN)` was not making a fullscreen window.
##### Steps to reproduce
The following lines will not produce a fullscreen window:
```.cpp
namedWindow(name, WINDOW_FULLSCREEN);
setWindowProperty(name, WND_PROP_FULLSCREEN, WINDOW_FULLSCREEN);
imshow(name, frame);
```
The following lines *will* produce the desired fullscreen result:
```.cpp
namedWindow(name, WINDOW_NORMAL);
setWindowProperty(name, WND_PROP_FULLSCREEN, WINDOW_FULLSCREEN);
imshow(name, frame);
``` | category: highgui-gui,category: documentation | low | Minor |
284,785,041 | vue | Required inputs in child components are highlighted as invalid on render in Firefox. | ### Version
2.5.13
### Reproduction link
[https://jsfiddle.net/50wL7mdz/85928/](https://jsfiddle.net/50wL7mdz/85928/)
### Steps to reproduce
No additional steps required. Opening the JSFiddle in Firefox should automatically cause the issue to display.
### What is expected?
Required inputs that are rendered in a child component without a valid default value should display normally.
### What is actually happening?
Unselected required inputs are being highlighted as though a submit attempt was made despite no user action being taken.
<!-- generated by vue-issues. DO NOT REMOVE --> | browser quirks | low | Minor |
284,789,111 | kubernetes | Remove direct write access to this repo | EDIT: updated as of 2019-07-19
We now have sufficient automation in place to address most if not all of the use cases that would require direct write access to this repo:
- https://prow.k8s.io/tide: we allow people without direct write access to merge via /lgtm, /approve, and OWNERS files
- https://go.k8s.io/bot-commands: we allow people who are org members to assign, triage, and manage issues and their labels via /commands
The following teams have read access. I would like to remove these from the repo, they are effectively no-ops. You need to be an org member to join a team, and everyone in the org can read this repo.
- ~@kubernetes/goog-gke~ - Read access removed
- ~@kubernetes/kubernetes-reviewers~ - team deleted
- ~@kubernetes/api-reviewers~ - Read access removed
- ~@kubernetes/api-approvers~ - Read access removed
The following teams have write access:
- @kubernetes/kubernetes-maintainers - I would like to drastically reduce membership
- ~@kubernetes/bots~ - Access has been removed
- ~@kubernetes/kubernetes-pm~ - Team deprecated and removed
The following teams have admin access
- @kubernetes/kubernetes-admins
- @kubernetes/kubernetes-release-managers - I would like to reduce membership of this team to _active_ release and patch release managers
- ~@kubernetes/kubernetes-build-cops~ - Team deprecated and removed
- @kubernetes/steering-committee
What use cases must be implemented before we can kick (mostly) everybody out?
Summarizing from kubernetes/contrib#1908 and below, here is what has been addressed:
- ~component/foo labels~ - there are no `component/foo` labels in this repo
- ~team/foo labels?~ - there are no `team/foo` labels anymore
- ~add/remove milestones to issues/PR's~ - we now have `/milestone` commands
- ~wiki~ - we disabled the wiki (https://github.com/kubernetes/kubernetes/issues/57689)
- ~edited issue/PR titles~ - we now have a `/retitle` command (https://github.com/kubernetes/test-infra/pull/12945)
Here is what's left:
- editing other people's issues/PR's (descriptions)
- it's been a while since we've checked in on whether this is still a common practice; I suspect this is mostly in service of editing release notes
- improving our release note process or automation could obviate the need for some of this
- creating/editing milestones
- we lack automation to do this, at this point I am unconvinced this is a blocker | priority/backlog,kind/cleanup,sig/contributor-experience,lifecycle/frozen | medium | Critical |
284,802,288 | go | net/http: Request not cancelled when body is not read | ### What version of Go are you using (`go version`)?
go version go1.9.2 darwin/amd64
### Does this issue reproduce with the latest release?
yes
### What operating system and processor architecture are you using (`go env`)?
GOARCH="amd64"
GOBIN="/Users/sheerun/go/bin"
GOEXE=""
GOHOSTARCH="amd64"
GOHOSTOS="darwin"
GOOS="darwin"
GOPATH="/Users/sheerun/go"
GORACE=""
GOROOT="/usr/local/Cellar/go/1.9.2/libexec"
GOTOOLDIR="/usr/local/Cellar/go/1.9.2/libexec/pkg/tool/darwin_amd64"
GCCGO="gccgo"
CC="clang"
GOGCCFLAGS="-fPIC -m64 -pthread -fno-caret-diagnostics -Qunused-arguments -fmessage-length=0 -fdebug-prefix-map=/var/folders/22/xz6_9gpx3jggts_8j68_25g80000gn/T/go-build547040878=/tmp/go-build -gno-record-gcc-switches -fno-common"
CXX="clang++"
CGO_ENABLED="1"
CGO_CFLAGS="-g -O2"
CGO_CPPFLAGS=""
CGO_CXXFLAGS="-g -O2"
CGO_FFLAGS="-g -O2"
CGO_LDFLAGS="-g -O2"
PKG_CONFIG="pkg-config"
### What did you do?
I wanted proper go-lang server that detects when request is cancelled:
```go
package main
import (
"fmt"
"net/http"
"time"
)
func Handler(w http.ResponseWriter, req *http.Request) {
go func() {
<-req.Context().Done()
fmt.Println("done")
}()
time.Sleep(10 * time.Second)
}
func main() {
http.ListenAndServe(":8080", http.HandlerFunc(Handler))
}
```
### What did you expect to see?
Request is immediately cancelled for following calls:
```
curl http://localhost:8080/hello # then Ctrl+C
curl http://localhost:8080/hello -X POST -d 'asdfa' # then Ctrl+C
curl http://localhost:8080/hello -X PUT -d 'foobar' # then Ctrl+C
```
### What did you see instead?
Request is not cancelled for POST and PUT with body.
Side note: when POST of PUT is sent without body, the request is cancelled immediately as expected
```
curl http://localhost:8080/hello -X POST # then Ctrl+C
curl http://localhost:8080/hello -X PUT # then Ctrl+C
``` | Thinking,help wanted | medium | Critical |
284,817,671 | TypeScript | strictPropertyInitialization with subclasses that refine the types of properties | <!-- Please try to reproduce the issue with `typescript@next`. It may have already been fixed. -->
**TypeScript Version:** 2.7.0-dev.20171226
**Code**
```ts
class A {
x: {y: string} = {y: 'y'};
}
class B extends A {
x: {y: string, z?: number};
}
```
**Expected behavior:**
I expected this not to warn because `A#x` is assignable to `B#x`.
**Actual behavior:**
> index.ts(6,3): error TS2564: Property 'x' has no initializer and is not definitely assigned in the constructor.
| Suggestion,In Discussion | low | Critical |
284,847,655 | opencv | Kinect Sensor not working with OpenCV3 and OpenNI2 | ##### System information (version)
- OpenCV => 3.3
- Operating System / Platform => Ubuntu 16 ARM64 (aarch64)
- Compiler => g++ (Ubuntu/Linaro 5.4.0-6ubuntu1~16.04.5) 5.4.0 20160609
##### Detailed description
My Jetson TX1 ARM64 ubuntu 16 system is running libfreenect2 (Protonect) and OpenNi2 (NiViewer2) successfully. I read Opencv 3 has some additional requirements (https://docs.opencv.org/3.0-beta/doc/user_guide/ug_highgui.html) for working with kinect sensor data like PrimeSensor Module (OPENNI_PRIME_SENSOR_MODULE_BIN_DIR). Opencv was built with -DWITH_OPENNI2=ON and compiles fine, but I run into the following error when I ran cpp-example-openni_capture:
/home/nvidia/opencv/modules/videoio/src/cap_openni2.cpp:299: error: (-2) CvCapture_OpenNI2::CvCapture_OpenNI2 : Couldn't set color stream output mode: Stream setProperty(3) failed in function toggleStream
##### Steps to reproduce
Nvidia Jetson TX1 (hardware)
Xbox One Kinect Sensor (hardware)
Install libfreenect2
Install OpenNi2
Install OpenCV 3.3 with examples
Install Sensor Kinect (https://github.com/antoniodourado/SensorKinectTX1)
Run cpp-example-openni_capture opencv example
---------------------------------------------------------------------------------------------------------------------
```.sh
(testpy3) nvidia@tegra-ubuntu:~/opencv/build/bin$ ./cpp-example-openni_capture
Device opening ...
3583 INFO New log started on 2017-12-27 21:52:39
3615 INFO --- Filter Info --- Minimum Severity: VERBOSE
3672 VERBOSE No override device in configuration file
3710 VERBOSE Configuration has been read from '/etc/openni2/OpenNI.ini'
3731 VERBOSE OpenNI 2.2.0 (Build 33)-Linux-generic (Jan 24 2016 23:21:00)
3747 VERBOSE Extending the driver path by '/usr/lib/OpenNI2/Drivers', as configured in file '/etc/openni2/OpenNI.ini'
3766 VERBOSE Using '/usr/lib/OpenNI2/Drivers' as driver path
3773 VERBOSE Looking for drivers in drivers repository '/usr/lib/OpenNI2/Drivers'
15738 INFO New log started on 2017-12-27 21:52:39
15764 INFO --- Filter Info --- Minimum Severity: VERBOSE
15778 VERBOSE Initializing USB...
29881 INFO USB is initialized.
37006 INFO New log started on 2017-12-27 21:52:39
37031 INFO --- Filter Info --- Minimum Severity: VERBOSE
37053 VERBOSE Initializing USB...
51784 INFO USB is initialized.
[Info] [Freenect2Impl] enumerating devices...
[Info] [Freenect2Impl] 14 usb devices connected
[Info] [Freenect2Impl] found valid Kinect v2 @2:6 with serial 015802564847
[Info] [Freenect2Impl] found 1 devices
187432 INFO Found device freenect2://0
187445 INFO Driver: register new uri: freenect2://0
187456 INFO Device connected: Microsoft Kinect (freenect2://0)
187476 INFO Device state changed: Microsoft Kinect (freenect2://0) to 0
187489 INFO Driver: register new uri: freenect2://0?depth-size=640x480
187498 INFO Device connected: Microsoft Kinect (freenect2://0?depth-size=640x480)
187509 INFO Device state changed: Microsoft Kinect (freenect2://0?depth-size=640x480) to 0
187519 INFO Driver: register new uri: freenect2://0?depth-size=512x424
187527 INFO Device connected: Microsoft Kinect (freenect2://0?depth-size=512x424)
187538 INFO Device state changed: Microsoft Kinect (freenect2://0?depth-size=512x424) to 0
188899 VERBOSE Trying to open device by URI '(NULL)'
188931 INFO deiveOpen: freenect2://0
188941 INFO Opening device freenect2://0
[Info] [Freenect2DeviceImpl] opening...
[Info] [Freenect2DeviceImpl] transfer pool sizes rgb: 2016384 ir: 608*33792
[Info] [Freenect2DeviceImpl] opened
437992 INFO Device: createStream(depth)
438115 INFO Freenect2Driver::Device: start()
[Info] [Freenect2DeviceImpl] starting...
[Info] [Freenect2DeviceImpl] submitting rgb transfers...
[Info] [Freenect2DeviceImpl] submitting depth transfers...
[Info] [Freenect2DeviceImpl] started
1051623 INFO Device: createStream(color)
OpenCV Error: Unspecified error (CvCapture_OpenNI2::CvCapture_OpenNI2 : Couldn't set color stream output mode: Stream setProperty(3) failed
) in toggleStream, file /home/nvidia/opencv/modules/videoio/src/cap_openni2.cpp, line 299
1052240 INFO Device: destroyStream(color)
1052285 INFO Freenect2Driver::Device: stop()
[Info] [RgbPacketStreamParser] packetsize or sequence doesn't match!
[Info] [Freenect2DeviceImpl] stopping...
[Info] [Freenect2DeviceImpl] canceling rgb transfers...
[Info] [Freenect2DeviceImpl] canceling depth transfers...
[Info] [Freenect2DeviceImpl] stopped
2088603 INFO Device: destroyStream(depth)
2088646 INFO Closing device freenect2://0
2088653 INFO Freenect2Driver::Device: stop()
2088657 INFO Freenect2Driver::Device: close()
2088662 INFO Freenect2Driver::Device: stop()
[Info] [Freenect2DeviceImpl] closing...
[Info] [Freenect2DeviceImpl] releasing usb interfaces...
[Info] [Freenect2DeviceImpl] deallocating usb transfer pools...
[Info] [Freenect2DeviceImpl] closing usb device...
[Info] [Freenect2DeviceImpl] closed
VIDEOIO(cvCreateCameraCapture_OpenNI2(index)): raised OpenCV exception:
/home/nvidia/opencv/modules/videoio/src/cap_openni2.cpp:299: error: (-2) CvCapture_OpenNI2::CvCapture_OpenNI2 : Couldn't set color stream output mode: Stream setProperty(3) failed
in function toggleStream
CvCapture_OpenNI::CvCapture_OpenNI : Failed to enumerate production trees: Can't create any node of the requested type!
done.
Can not open a capture object.
``` | priority: low,category: videoio(camera) | low | Critical |
284,863,827 | opencv | Incorrect work with border cornerHarris and cornerMinEigenVal | ##### System information (version)
- OpenCV => 3.3.1
- Operating System / Platform => Ubuntu 16.04
- Compiler => gcc version 5.4.0
##### Detailed description
Function `cornerHarris` and `cornerMinEigenVal` works with border incorrect. Еspecially, it can be demonstrated on submat. At first, this algorithm call `Sobel`, and borders are interpolated as usual. So we have covariation matrix with size `src.width * src.height`. After that, used `boxFilter`, and borders interpolated one more time. But in submat case, we could calculete `Sobel` for border pixels and use it in Box. So in OpenCV case, if you separate matrix for submats, and calculates cornerHarris for full mat and submats, you will have different result for pixel on submats borders. | bug,category: imgproc | low | Minor |
284,896,980 | angular | Provide extension points in order to add custom behavior to the service worker file generated by the angular cli. | <!--
PLEASE HELP US PROCESS GITHUB ISSUES FASTER BY PROVIDING THE FOLLOWING INFORMATION.
ISSUES MISSING IMPORTANT INFORMATION MAY BE CLOSED WITHOUT INVESTIGATION.
-->
## I'm submitting a...
<!-- Check one of the following options with "x" -->
<pre><code>
[ ] Regression (a behavior that used to work and stopped working in a new release)
[ ] Bug report <!-- Please search GitHub for a similar issue or PR before submitting -->
[x] Feature request
[x] Documentation issue or request
[ ] Support request => Please do not submit support request here, instead see https://github.com/angular/angular/blob/master/CONTRIBUTING.md#question
</code></pre>
## Current behavior
<!-- Describe how the issue manifests. -->
As of now, I can either register the angular-generated service worker (i.e. `/ngsw-worker.js`) or register my own custom service worker (see documentation: https://angular.io/api/service-worker/ServiceWorkerModule#static-members)
## Expected behavior
<!-- Describe what the desired behavior would be. -->
One should be able to add custom behavior to angular-generated service workers (i.e. `/ngsw-worker.js`). For instance what if I want to add business logic to the service worker. It would be nice to provide extension points in order to achieve this and document them.
##
_**EDIT** (gkalpak): Other related issues/issues that could be fixed by this: #24387, #26413, #30195, #46984_
| feature,area: service-worker,feature: under consideration | medium | Critical |
284,959,452 | flutter | API doc clarification request: if a key is not provided for a widget, what is default key? | Trying to learn about keys, and I found https://docs.flutter.io/flutter/foundation/Key-class.html
One of my initial questions was: "what is the "default" key?" and "if I don't provide a key, what's the default?"
I also checked https://docs.flutter.io/flutter/widgets/Widget/key.html and didn't find an obvious answer. | framework,d: api docs,has reproducible steps,P2,found in release: 3.3,found in release: 3.5,team-framework,triaged-framework | low | Major |
284,960,285 | TypeScript | TS Proposal : "Interface incorrectly extends interface" - sub-interface method overload OR override ? | **TypeScript Version :** 2.6.2
## The problem
**Initial scenario**
I encountered the issue yesterday days as I was trying to add some features to one of my interface inheritance chain. The context of this example is just imagined : It is longer than needed to demonstrate the issue, but I wanted a full example to clearly show it needs to be addressed.
So, let's imagine a testing app that should instrument devices : the following is a potential model of the situation :
```ts
interface Stream
{
}
interface InputStream
{
}
interface OutputStream
{
}
// interface inheraitance is a great feature
interface IOStream extends InputStream, OutputStream
{
}
```
```ts
interface InterfaceLike
{
}
interface DeviceInterfaceLike extends InterfaceLike
{
init();
}
interface InputDeviceInterfaceLike extends DeviceInterfaceLike
{
input(); // this should word just like the `touch` command
// like if there's and error with the device, it will trigger exception
input(stream: InputStream);
}
```
```ts
interface ScreenDeviceInterfaceLike extends InputDeviceInterfaceLike
{
input(stream: InputStream, coordinates: {x: number, y: number});
}
interface TouchScreenDeviceInterfaceLike extends InputDeviceInterfaceLike
{
input(stream: IOStream, coordinates: {x: number, y: number});
}
```
**Expected behavior :**
It should just overload that method in the sub-interfaces.
**Actual behavior :**
```
error TS2430: Interface 'ScreenDeviceInterfaceLike' incorrectly extends interface 'InputDeviceInterfaceLike'.
Types of property 'input' are incompatible.
Type '(stream: InputStream, coordinates: { x: number; y: number; }) => any' is not assignable to type '{ (): any; (stream: InputStream): any; }'.
core.ts(73,11): error TS2430: Interface 'TouchScreenDeviceInterfaceLike' incorrectly extends interface 'InputDeviceInterfaceLike'.
Types of property 'input' are incompatible.
Type '(stream: IOStream, coordinates: { x: number; y: number; }) => any' is not assignable to type '{ (): any; (stream: InputStream): any; }'.
18:38:26 - Compilation complete. Watching for file changes.
```
It requires me to fully copy paste all method signatures in each interface.
```ts
interface ScreenDeviceInterfaceLike extends InputDeviceInterfaceLike
{
input(); // this should word just like the `touch` command
// like if there's and error with the device, it will trigger exception
input(stream: InputStream);
input(stream: InputStream, coordinates: {x: number, y: number});
}
interface TouchScreenDeviceInterfaceLike extends InputDeviceInterfaceLike
{
input(); // this should word just like the `touch` command
// like if there's and error with the device, it will trigger exception
input(stream: InputStream);
input(stream: IOStream, coordinates: {x: number, y: number});
}
```
When you're can have up to (why not) 15 overloads distributed through inheritance chain... It becomes, (yes, you said it) bulky.
-----
Now it is, that the compiler cannot yet figure out when to `override` the method signatures inherited from parents and when to `overload` them.
## The Proposal
My first thought was to introduce a new keyword just like `@ts-nocheck` and like.
### `@override`
```ts
interface InputDeviceInterfaceLike extends DeviceInterfaceLike
{
@override // to override all definitions from parents
init(istream: InputStream);
}
```
OR
```ts
interface InputDeviceInterfaceLike extends DeviceInterfaceLike
{
@override
{
init(); // to override only this (these) definitions : forget all previous definition but this (these)
}
init(istream: InputStream);
}
```
Then, the `TouchScreenDeviceInterfaceLike` interface would become :
```ts
interface TouchScreenDeviceInterfaceLike extends InputDeviceInterfaceLike
{
@override // it is now clear that the only way to instrument a touchscreen is to provide a `IOStream` and coordinates (`{x: number, y: number}`)
input(stream: IOStream, coordinates: {x: number, y: number});
}
```
### Furthermore, `@overload`
While the previous **`@override` is useful when the default behavior is oveloading**, we can still define the default behavior to **suppress all parent definition WHEN A METHOD IS REDEFINED in subinterface, except when `@overload` is used**. i.e
```ts
interface InputDeviceInterfaceLike extends DeviceInterfaceLike
{
// to override all definitions from parents
init(istream: InputStream);
}
```
OR
```ts
interface InputDeviceInterfaceLike extends DeviceInterfaceLike
{
@overload // to overload signatures from parents
init(istream: InputStream);
}
```
And `TouchScreenDeviceInterfaceLike` interface can become :
```ts
interface TouchScreenDeviceInterfaceLike extends InputDeviceInterfaceLike
{
@overload // to overload signatures from parents
{
init(istream: InputStream); // except this signature... So that it remains only the empty parameter and the below
}
input(stream: IOStream, coordinates: {x: number, y: number});
}
```
----
By thinking rigorously, I think the first step would be allow inheriting type to overload methods by default.
**= = = = = = UPDATE = = = = = =**
I now think it worth to add that by no mean, all through an interface inheritance chain should a method be completely wiped off (deleted, removed)... Not even by `@Override`! As this may (will) break the core concept of OOP inheritance.
The mechanism is to ensure that only some methods mood (signatures) are handled (accepted) at a given inheritance node.
In other words, if `@Override` suppresses all signatures, at least one method signature MUST be required. | Suggestion,Needs Proposal | medium | Critical |
284,965,010 | opencv | Build for OSX fails | Hi, the build_framework for osx is out of date as you stated earlier. Is there any workaround that you know of that can help me build OSX stuff without Cuda or perhaps another module?
The issue is that running the script that generates the OpenCV3.4 framework for use on OSX is not working, in case others reading are wondering. | category: build/install,platform: ios/osx,incomplete | low | Minor |
284,992,453 | go | cmd/dist: detect FPUless ARMv7 platforms? | Hi,
I'm trying to build Go on an FPUless ARMv7 platform:
```
[root@unknown src]$ cat /proc/cpuinfo
Processor : ARMv7 Processor rev 0 (v7l)
processor : 0
BogoMIPS : 1599.07
processor : 1
BogoMIPS : 1595.80
Features : swp half thumb fastmult edsp
CPU implementer : 0x41
CPU architecture: 7
CPU variant : 0x3
CPU part : 0xc09
CPU revision : 0
Hardware : Northstar Prototype
Revision : 0000
Serial : 0000000000000000
```
`gccgo-go` from GCC-7.2.0 is used as the bootstrap toolchain:
```
[root@unknown src]$ go version
go version go1.8.1 arm-buildroot-linux-uclibcgnueabi-gccgo (Buildroot 2017.08-g56f1bab-dirty) 7.2.0 linux/arm
[root@unknown src]$ go env
GOARCH="arm"
GOBIN=""
GOEXE=""
GOHOSTARCH="arm"
GOHOSTOS="linux"
GOOS="linux"
GOPATH="/root/go"
GORACE=""
GOROOT="/opt"
GOTOOLDIR="/opt/libexec/gcc/arm-buildroot-linux-uclibcgnueabi/7.2.0"
GCCGO="/opt/bin/gccgo"
GOARM=""
CC="/opt/bin/gcc"
GOGCCFLAGS="-fPIC -marm -pthread -fmessage-length=0 -fdebug-prefix-map=/tmp/go-build935653095=/tmp/go-build -gno-record-gcc-switches"
CXX="/opt/bin/g++"
CGO_ENABLED="1"
PKG_CONFIG="pkg-config"
CGO_CFLAGS="-g -O2"
CGO_CPPFLAGS=""
CGO_CXXFLAGS="-g -O2"
CGO_FFLAGS="-g -O2"
CGO_LDFLAGS="-g -O2"
```
I tried both `go1.9.2` and `master` branch. Here's build log for the latter:
```
[root@unknown src]$ GOROOT_FINAL=/opt/local CGO_ENABLED=1 GOROOT_BOOTSTRAP=/opt ./make.bash
Building Go cmd/dist using /opt.
Building Go toolchain1 using /opt.
Building Go bootstrap cmd/go (go_bootstrap) using Go toolchain1.
Building Go toolchain2 using go_bootstrap and Go toolchain1.
go tool dist: FAILED: /opt/src/golang/pkg/tool/linux_arm/go_bootstrap install -gcflags=all= -ldflags=all= -i cmd/asm cmd/cgo cmd/compile cmd/link: signal: Illegal instruction
```
And here's a `gdb` log:
```
[root@unknown src]$ gdb /opt/src/golang/pkg/tool/linux_arm/go_bootstrap
GNU gdb (GDB) 8.0.1
Copyright (C) 2017 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law. Type "show copying"
and "show warranty" for details.
This GDB was configured as "arm-linux".
Type "show configuration" for configuration details.
For bug reporting instructions, please see:
<http://www.gnu.org/software/gdb/bugs/>.
Find the GDB manual and other documentation resources online at:
<http://www.gnu.org/software/gdb/documentation/>.
For help, type "help".
Type "apropos word" to search for commands related to "word"...
Reading symbols from /opt/src/golang/pkg/tool/linux_arm/go_bootstrap...done.
warning: Unsupported auto-load script at offset 0 in section .debug_gdb_scripts
of file /opt/src/golang/pkg/tool/linux_arm/go_bootstrap.
Use `info auto-load python-scripts [REGEXP]' to list them.
(gdb) run
Starting program: /opt/src/golang/pkg/tool/linux_arm/go_bootstrap
Program received signal SIGILL, Illegal instruction.
runtime.check () at /opt/local/src/runtime/runtime1.go:146
146 i, i1 float32
(gdb) bt full
#0 runtime.check () at /opt/local/src/runtime/runtime1.go:146
e = 0
i = 0
j = 0
k = 0x0
m = "\000\000\000"
z = 0
#1 0x00062e94 in runtime.rt0_go () at /opt/local/src/runtime/asm_arm.s:149
No locals.
#2 0x00000000 in ?? ()
No symbol table info available.
Backtrace stopped: previous frame identical to this frame (corrupt stack?)
```
The line `/opt/local/src/runtime/asm_arm.s:149` is:
```
BL runtime·check(SB)
```
In `go1.9.2` it crashes on `BL runtime·check(SB)` as well.
I should also note that uClibc-ng 1.0.27 is used as the libc.
Regards,
Alex | help wanted,NeedsFix | low | Critical |
284,994,686 | pytorch | Cache CuDNN benchmark selection, turn it on by default, use it across PyTorch runs | CuDNN benchmark mode can have some beefy speed increases for our cuDNN users, but we don't turn it on by default because the benchmarking process takes a long time. One possibility to alleviate this so that we can turn it on by default is to cache the benchmark selection (keyed on PyTorch version, cuDNN version, device ID, and all of the input parameters) in, say, `$HOME/.pytorch`, and then reuse benchmark information on subsequent runs.
The primary hazard is the introduction of state, so it's very important to not accidentally state-dependence (which will lead to hard to debug performance regressions and problems.)
Related: https://github.com/pytorch/pytorch/issues/3667
cc @csarofeen @ptrblck | module: cudnn,triaged | low | Critical |
284,994,930 | go | cmd/gofmt: -r 'bar(x) -> bar(x)' mistakenly replaces a comment with a newline in closures inside function | ### What version of Go are you using (`go version`)?
go version go1.9.2 linux/amd64
### What did you do?
`$ gofmt -r 'bar(x) -> bar(x)' /tmp/test.go`
/tmp/test.go is:
```
package main
func main() {
// Call foo().
foo()
bar(func() {
// Call foo().
foo()
})
}
```
### What did you expect to see?
Same contents
### What did you see instead?
```
package main
func main() {
// Call foo().
foo()
bar(func() {
foo()
})
}
```
Notice the comment in the closure passed to bar was replaced with an empty line, and an empty line was added after the call to bar.
The identity rewrite rule is for simplicity, this happens with more useful rewrites that act on bar too. | NeedsFix | low | Minor |
285,016,605 | go | cmd/link: nicer error when trying to link an invalid .syso file built on another OS | ### What version of Go are you using (`go version`)?
```
go version go1.9.2 linux/amd64
```
### Does this issue reproduce with the latest release?
Yes
### What operating system and processor architecture are you using (`go env`)?
```
$ go env
GOARCH="amd64"
GOBIN=""
GOEXE=""
GOHOSTARCH="amd64"
GOHOSTOS="linux"
GOOS="linux"
GOPATH="/home/litarvan/go"
GORACE=""
GOROOT="/usr/lib/go"
GOTOOLDIR="/usr/lib/go/pkg/tool/linux_amd64"
GCCGO="gccgo"
CC="gcc"
GOGCCFLAGS="-fPIC -m64 -pthread -fmessage-length=0 -fdebug-prefix-map=/tmp/go-build281376345=/tmp/go-build -gno-record-gcc-switches"
CXX="g++"
CGO_ENABLED="1"
CGO_CFLAGS="-g -O2"
CGO_CPPFLAGS=""
CGO_CXXFLAGS="-g -O2"
CGO_FFLAGS="-g -O2"
CGO_LDFLAGS="-g -O2"
PKG_CONFIG="pkg-config"
```
### What did you do?
I did, on linux, `go build .` on a project firstly created on Windows, containing a `main.go` file but also a `resource.syso` file generated with winres, that contains my icon file for the windows executable. The build was from linux this time, targeting linux, so i didn't pay attention to the .syso file.
### What did you expect to see?
A working build
### What did you see instead?
`go build .` dropped this error :
```
/usr/lib/go/pkg/tool/linux_amd64/link: running gcc failed: exit status 1
/tmp/go-link-918954239/000000.o: file not recognized: File format not recognized
collect2: error: ld returned exit status 1
```
It took me like 1 hour to think that `go build` was trying to include the .syso file even when i was on linux, and was crashing the build because of this, i didn't really pay attention to the command as it was issued by a script.
go build should ignore .syso file on systems other than windows, or (i don't know if .syso is a windows-only extension) display a more readable error | NeedsFix,compiler/runtime | low | Critical |
285,019,665 | pytorch | Installation Optimise For Chinese Users Who Behind the Wall | As you know, we are just can not access all Google etc access behind the f*k great wall. The pytorch.org wesite can not load properly when click pip and cuda chips. Just not responding, I think the site using some access which block in China. Just for suggestion, write more install command in Github README.md or anywhere else.
cc @jlin27 | module: docs,triaged | low | Major |
285,071,995 | pytorch | DistributedDataParallel doesn't converge well when using MPI | ## Originally posted by @alsrgv:
I got Gloo working with 4 nodes with 4 GPUs each, but it fails on 32 nodes. I tried MPI backend with `DistributedDataParallel` and found that it doesn't converge very well.
I scaled back experiment to 4 nodes with 4 GPUs each and observed that MPI backend with `DistributedDataParallel` indeed converges much worse with the same batch size & lr. On the chart below, gray line is Gloo and green line is MPI.
<img width="714" alt="screen shot 2017-12-27 at 4 03 34 pm" src="https://user-images.githubusercontent.com/16640218/34396423-934bfc4e-eb1f-11e7-8fca-8bd3a9ed8b33.png">
I was finally able to run large-scale experiment with doing a simple gradient averaging instead of using `DistributedDataParallel` functionality:
```
def average_gradients(model):
size = float(dist.get_world_size())
for param in model.parameters():
dist.all_reduce(param.grad.data, op=dist.reduce_op.SUM)
param.grad.data /= size
```
I did two more runs - one with `DataParallel` and one worker per machine (blue line), and another is with one worker per GPU (orange line).
<img width="707" alt="screen shot 2017-12-27 at 4 21 27 pm" src="https://user-images.githubusercontent.com/16640218/34396602-059783b6-eb22-11e7-9ba1-079bcbbe5a04.png">
Unfortunately, it's slow. Any idea how to make `DistributedDataParallel` work correctly with MPI?
cc @pietern @mrshenli @pritamdamania87 @zhaojuanmao @satgera @rohan-varma @gqchen @aazzolini @xush6528 @osalpekar @jiayisuse @agolynski | oncall: distributed,triaged | low | Major |
285,146,868 | youtube-dl | Unable to download from http://members.bangbros.com | ## Please follow the guide below
- You will be asked some questions and requested to provide some information, please read them **carefully** and answer honestly
- Put an `x` into all the boxes [ ] relevant to your *issue* (like this: `[x]`)
- Use the *Preview* tab to see what your issue will actually look like
---
### Make sure you are using the *latest* version: run `youtube-dl --version` and ensure your version is *2017.12.28*. If it's not, read [this FAQ entry](https://github.com/rg3/youtube-dl/blob/master/README.md#how-do-i-update-youtube-dl) and update. Issues with outdated version will be rejected.
- [ ] I've **verified** and **I assure** that I'm running youtube-dl **2017.12.28**
### Before submitting an *issue* make sure you have:
- [ ] At least skimmed through the [README](https://github.com/rg3/youtube-dl/blob/master/README.md), **most notably** the [FAQ](https://github.com/rg3/youtube-dl#faq) and [BUGS](https://github.com/rg3/youtube-dl#bugs) sections
- [ ] [Searched](https://github.com/rg3/youtube-dl/search?type=Issues) the bugtracker for similar issues including closed ones
### What is the purpose of your *issue*?
- [ ] Bug report (encountered problems with youtube-dl)
- [ ] Site support request (request for adding support for a new site)
- [ ] Feature request (request for a new functionality)
- [ ] Question
- [ ] Other
---
### The following sections concretize particular purposed issues, you can erase any section (the contents between triple ---) not applicable to your *issue*
---
### If the purpose of this *issue* is a *bug report*, *site support request* or you are not completely sure provide the full verbose output as follows:
Add the `-v` flag to **your command line** you run youtube-dl with (`youtube-dl -v <your command line>`), copy the **whole** output and insert it here. It should look similar to one below (replace it with **your** log inserted between triple ```):
```
[debug] System config: []
[debug] User config: []
[debug] Command-line args: [u'-v', u'http://www.youtube.com/watch?v=BaW_jenozKcj']
[debug] Encodings: locale cp1251, fs mbcs, out cp866, pref cp1251
[debug] youtube-dl version 2017.12.28
[debug] Python version 2.7.11 - Windows-2003Server-5.2.3790-SP2
[debug] exe versions: ffmpeg N-75573-g1d0487f, ffprobe N-75573-g1d0487f, rtmpdump 2.4
[debug] Proxy map: {}
...
<end of log>
```
---
### If the purpose of this *issue* is a *site support request* please provide all kinds of example URLs support for which should be included (replace following example URLs by **yours**):
- Single video: https://www.youtube.com/watch?v=BaW_jenozKc
- Single video: https://youtu.be/BaW_jenozKc
- Playlist: https://www.youtube.com/playlist?list=PL4lCao7KL_QFVb7Iudeipvc2BCavECqzc
Note that **youtube-dl does not support sites dedicated to [copyright infringement](https://github.com/rg3/youtube-dl#can-you-add-support-for-this-anime-video-site-or-site-which-shows-current-movies-for-free)**. In order for site support request to be accepted all provided example URLs should not violate any copyrights.
---
### Description of your *issue*, suggested solution and other information
Explanation of your *issue* in arbitrary form goes here. Please make sure the [description is worded well enough to be understood](https://github.com/rg3/youtube-dl#is-the-description-of-the-issue-itself-sufficient). Provide as much context and examples as possible.
If work on your *issue* requires account credentials please provide them or explain how one can obtain them.
When I try to download in bulk using Youtube-dl, I get the following:
./youtube-dl --cookies cookies.txt -F http://members.bangbros.com/product/495
[generic] 495: Requesting header
[redirect] Following redirect to http://members.bangbros.com/login
[generic] login: Requesting header
WARNING: Falling back on generic information extractor.
[generic] login: Downloading webpage
[generic] login: Extracting information
ERROR: Unsupported URL: http://members.bangbros.com/login
./youtube-dl -U
youtube-dl is up-to-date (2017.12.28)
./youtube-dl --cookies cookies.txt -F http://members.bangbros.com/product/495
[generic] 495: Requesting header
[redirect] Following redirect to http://members.bangbros.com/login
[generic] login: Requesting header
WARNING: Falling back on generic information extractor.
[generic] login: Downloading webpage
[generic] login: Extracting information
ERROR: Unsupported URL: http://members.bangbros.com/login
I get the same error message when I try individual videos too. | account-needed,nsfw | low | Critical |
285,147,334 | create-react-app | Add WebWorker Support | Add ability to load WebWorkers (and SharedWorkers?) via the react-scripts. My use-case is for using latency (and debug) sensitive protocols over WebSocket which may get dumped by the server if no activity occurs.
There has been a lot of discussion regarding this on #1277 but no definitive answer and now closed. Also, I'm a server-side dev so JavaScript PR is not my forte, and neither is ejecting the scripts as that looks scary. | issue: proposal | high | Critical |
285,153,379 | opencv | Inconsistent return type from CascadeClassifier.detectMultiScale in Python | ##### System information (version)
- OpenCV => 3.3.1
- Operating System / Platform => Windows 10 64 Bit
- Compiler =>Python 3.6.3
##### Detailed description
in CascadeClassifier.detectMultiScale, I expected the library to return an empty ndarray when there is no detection.
The library returns a single empty tuple when there is NO detection, and a numpy.ndarray when there are detections.
| category: python bindings,category: objdetect | low | Minor |
285,212,510 | react | Consider a more specific warning for key={undefined} | Proposed in [this comment](https://dev.to/k1sul1/comment/1o68):
>I had changed the casing of "ID" in the response, but forgot to commit it aaaaaand I ended up with it happening.
>Basically I was doing key={undefined}. Could React warn user when this happens, something like "Looks like you tried to supply a key, but the value supplied is undefined. Check the render..." and so on?
I think it might make sense to give a more specific warning in this case. Open to suggestions about specific wording and in which case it would be used. | Type: Enhancement,Component: Core Utilities,React Core Team | low | Major |
285,213,763 | ant-design | Expand all header | ### What problem does this feature solve?
It feels natural to have an expand all button alongside with `onExpandAll` property, the same way it's done for selection column. It can be achieved using existing api, but UI would be much cleaner with header button.
And perhaps it makes sense to group expandable options too. (see proposal api).
p.s. Thanks for the amazing product!
### What does the proposed API look like?
```
<Table
...
rowExpansion={{
expandedRowKeys: <...>,
expandedRowRender: <...>,
...
onExpandAll: <..>,
}}
/>
```
<!-- generated by ant-design-issue-helper. DO NOT REMOVE --> | 🗣 Discussion,💡 Feature Request,Inactive,IssueHuntFest | medium | Major |
285,221,198 | youtube-dl | [Request] Add Cablevision (Optimum) to --ap-mso | ## Please follow the guide below
- You will be asked some questions and requested to provide some information, please read them **carefully** and answer honestly
- Put an `x` into all the boxes [ ] relevant to your *issue* (like this: `[x]`)
- Use the *Preview* tab to see what your issue will actually look like
---
### Make sure you are using the *latest* version: run `youtube-dl --version` and ensure your version is *2017.12.28*. If it's not, read [this FAQ entry](https://github.com/rg3/youtube-dl/blob/master/README.md#how-do-i-update-youtube-dl) and update. Issues with outdated version will be rejected.
- [ x] I've **verified** and **I assure** that I'm running youtube-dl **2017.12.28**
### Before submitting an *issue* make sure you have:
- [ x] At least skimmed through the [README](https://github.com/rg3/youtube-dl/blob/master/README.md), **most notably** the [FAQ](https://github.com/rg3/youtube-dl#faq) and [BUGS](https://github.com/rg3/youtube-dl#bugs) sections
- [x ] [Searched](https://github.com/rg3/youtube-dl/search?type=Issues) the bugtracker for similar issues including closed ones
### What is the purpose of your *issue*?
- [ ] Bug report (encountered problems with youtube-dl)
- [ ] Site support request (request for adding support for a new site)
- [ x] Feature request (request for a new functionality)
- [x ] Question
- [ ] Other
---
### The following sections concretize particular purposed issues, you can erase any section (the contents between triple ---) not applicable to your *issue*
---
---
I have Optimum (Cablevision in NYC) but it is currently not an option in the --ap-mso list.
Sample url: http://www.wetv.com/shows/mama-june-from-not-to-hot/full-episode/season-01/the-confrontation
Looking at the source code, 'Optimum' is labeled as 'Cablevision'. I attempted to add the following but i am not able to successfully authenticate.
-- adobepass.py --
,
'Cablevision': {
'name': 'Cablevision',
'username_field': 'IDToken1',
'password_field': 'IDToken2',
}
Looking at Optimum's login page, it looks like there is 3 hidden forms and the mso goes for the first one which is for hte username part. I added the following in an attempt to get it to login
def post_form_cablevision(form_page_res, note, data={}):
form_page, urlh = form_page_res
post_url = self._html_search_regex(r'<form name[^>]+action=(["\'])(?P<url>.+?)\1', form_page, 'post url', group='url')
if not re.match(r'https?://', post_url):
post_url = compat_urlparse.urljoin(urlh.geturl(), post_url)
form_data = self._hidden_inputs(form_page)
form_data.update(data)
return self._download_webpage_handle(
post_url, video_id, note, data=urlencode_postdata(form_data), headers={
'Content-Type': 'application/x-www-form-urlencoded',
})
After the if mso_id == 'Comcast_SSO': section, I added
elif mso_id == 'Cablevision':
# This will be for Cablevision login
provider_redirect_page, urlh = provider_redirect_page_res
provider_refresh_redirect_url = extract_redirect_url( provider_redirect_page, url=urlh.geturl() )
if provider_refresh_redirect_url:
provider_redirect_page_res = self._download_webpage_handle( provider_refresh_redirect_url, video_id, 'Downloading Provider Redirect Page (meta refresh)')
provider_login_page_res = post_form( provider_redirect_page_res, self._DOWNLOADING_LOGIN_PAGE)
html_code = provider_login_page_res[0].split('\n')
for line in html_code:
if 'name="goto" value="' in line:
goto = line.strip().replace('<input type="hidden" name="goto" value="','').replace('">','')
if 'SunQueryParamsString' in line:
sunquery = line.strip().replace('<input type="hidden" name="SunQueryParamsString" value="','').replace('">','')
mvpd_confirm_page_res = post_form_cablevision(provider_login_page_res, 'Logging in', {
mso_info.get('username_field', 'username'): username,
mso_info.get('password_field', 'password'): password,
'goto': goto,
'SunQueryParamString': sunquery,
'encoded': 'true',
'gx_charset': 'UTF-8',
'IDButton': '',
})
When It tries to log in, i get a 401 - HTTP Error 401: Unauthorized -- Authentication Exception due to incorrect handler.
If i leave it with the original post_form then I get a HTTP Error 500: Internal Server Error
My apologies if I am rambling.
--- | tv-provider-account-needed | low | Critical |
285,222,401 | rust | Difference in trait object default bounds for early/late bound lifetimes | It looks like early- and late-bound lifetimes behave differently in default object bounds:
```rust
trait A<'a>: 'a {}
pub fn foo<'a>(x: Box<A<'a>>) -> Box<A<'a> + 'static> // Infers 'static
{
x
}
pub fn bar<'a>(x: Box<A<'a> + 'a>) -> Box<A<'a>> // Infers 'a
where 'a: 'a
{
x
}
``` | A-lifetimes,C-bug,T-types,A-trait-objects | low | Minor |
285,248,445 | flutter | Building on Ubuntu Xenial requires installing gcc-multilib and g++-multilib | ...and uninstalling `g*-arm-linux-gnueabihf` which got installed by our tools.
At least, that seems to have been the problem. I haven't tried to figure out exactly what part of `install-build-deps.sh` needs updating, since I'm not super familiar with the relevant packages and their interdependencies. | team,engine,platform-linux,P3,team-engine,triaged-engine | low | Minor |
285,249,874 | flutter | Need more error handling to handle connecting to an Android device that hasn't authorized the host | ```
ianh@burmese:~$ flutter devices
1 connected device:
Error retrieving device properties for ro.product.cpu.abi:
error: insufficient permissions for device
???????????? • ???????????? • android-arm • Android null (API null)
ianh@burmese:~$
``` | platform-android,tool,a: first hour,P2,team-android,triaged-android | low | Critical |
285,256,423 | vscode | Suggestion: Use RegEx classes in "editor.wordSeparators" settings | It looks like the settings string for "editor.wordSeparators" contains (almost) any punctuation character.
Wouldn't it make more sense to introduce the ability to utilize [regular expression character classes](https://docs.microsoft.com/en-us/dotnet/standard/base-types/character-classes-in-regular-expressions) for this purpose? | feature-request,editor-core | low | Minor |
285,286,917 | opencv | does cuda::cvtColor support the format of 'cv::COLOR_YUV2BGR_NV21'? | the following code will give an error:
cv::cuda::cvtColor(g_src, g_dst, cv::COLOR_YUV2BGR_NV21); | feature,priority: low,category: gpu/cuda (contrib) | low | Critical |
285,289,298 | go | x/build: we should run at least one windows builder as each of Administrator and non-Administrator | Some tests do not run unless they are run by Administrator.
Compare normal user runs:
```
c:\>go test -v -short os | grep -i skip
--- SKIP: TestStatError (0.00s)
testenv.go:189: skipping test: cannot make symlinks on windows/amd64: you don't have enough privileges to create symlinks
--- SKIP: TestReaddirNValues (0.00s)
os_test.go:515: test.short; skipping
--- SKIP: TestReaddirStatFailures (0.00s)
os_test.go:601: skipping test on windows
--- SKIP: TestSymlink (0.00s)
testenv.go:189: skipping test: cannot make symlinks on windows/amd64: you don't have enough privileges to create symlinks
--- SKIP: TestLongSymlink (0.00s)
testenv.go:189: skipping test: cannot make symlinks on windows/amd64: you don't have enough privileges to create symlinks
--- SKIP: TestLargeWriteToConsole (0.00s)
os_test.go:1817: skipping console-flooding test; enable with -large_write
--- SKIP: TestStatRelativeSymlink (0.00s)
testenv.go:189: skipping test: cannot make symlinks on windows/amd64: you don't have enough privileges to create symlinks
--- SKIP: TestSleep (0.00s)
os_test.go:2072: Skipping in short mode
--- SKIP: TestRemoveAllRace (0.00s)
os_test.go:2182: skipping on windows
--- SKIP: TestPipeThreads (0.00s)
os_test.go:2217: skipping on Windows; issue 19098
--- SKIP: TestDirectorySymbolicLink (0.03s)
os_windows_test.go:397: skipping some tests, could not enable "SeCreateSymbolicLinkPrivilege": Not all privileges or groups referenced are assigned to the caller.
--- SKIP: TestNetworkSymbolicLink (0.00s)
testenv.go:189: skipping test: cannot make symlinks on windows/amd64: you don't have enough privileges to create symlinks
--- SKIP: TestShareNotExistError (0.00s)
os_windows_test.go:547: slow test that uses network; skipping
--- SKIP: TestStatSymlinkLoop (0.00s)
testenv.go:189: skipping test: cannot make symlinks on windows/amd64: you don't have enough privileges to create symlinks
--- SKIP: TestRemoveAllLarge (0.00s)
path_test.go:176: skipping in short mode
--- SKIP: TestMkdirAllWithSymlink (0.00s)
testenv.go:189: skipping test: cannot make symlinks on windows/amd64: you don't have enough privileges to create symlinks
--- SKIP: TestMkdirAllAtSlash (0.00s)
path_test.go:234: skipping on windows
c:\>
```
to Administrator runs:
```
c:\>go test -v -short os | grep -i skip
--- SKIP: TestReaddirNValues (0.00s)
os_test.go:515: test.short; skipping
--- SKIP: TestReaddirStatFailures (0.00s)
os_test.go:601: skipping test on windows
--- SKIP: TestLargeWriteToConsole (0.00s)
os_test.go:1817: skipping console-flooding test; enable with -large_write
--- SKIP: TestSleep (0.00s)
os_test.go:2072: Skipping in short mode
--- SKIP: TestRemoveAllRace (0.00s)
os_test.go:2182: skipping on windows
--- SKIP: TestPipeThreads (0.00s)
os_test.go:2217: skipping on Windows; issue 19098
--- SKIP: TestShareNotExistError (0.00s)
os_windows_test.go:547: slow test that uses network; skipping
--- SKIP: TestRemoveAllLarge (0.00s)
path_test.go:176: skipping in short mode
--- SKIP: TestMkdirAllAtSlash (0.00s)
path_test.go:234: skipping on windows
c:\>
```
We have 5 windows builders at this moment. Perhaps we could make some run as Administrator to cover more tests.
Thank you.
Alex | OS-Windows,Builders | low | Critical |
285,318,302 | nvm | NVM D | <!-- Thank you for being interested in nvm! Please help us by filling out the following form if you‘re having trouble. If you have a feature request, or some other question, please feel free to clear out the form. Thanks! -->
- Operating system and version:
OSX 10.11.6
- `nvm debug` output:
No output from that command. It just disappeared back to the command prompt.
<!-- do not delete the following blank line -->
```sh
```
sh-3.2
- `nvm ls` output:
<details>
<!-- do not delete the following blank line -->
```sh
v9.3.0
default -> v0.10.32 (-> N/A)
node -> stable (-> N/A) (default)
unstable -> 9.3 (-> v9.3.0) (default)
iojs -> iojs- (-> N/A) (default)
```
</details>
- How did you install `nvm`? (e.g. install script in readme, homebrew):
install script
- What steps did you perform?
```sh
Als-iMac:.nvm alspohn$ nvm --version
0.25.1
Als-iMac:.nvm alspohn$ nvm use 0.25
N/A: version "v0.25" is not yet installed
```
- What happened?
It shows that -.25.1 is installed, but then says it isn't when I attempt to use it.
- What did you expect to happen?
I expected it to use 0.25.1
- Is there anything in any of your profile files (`.bashrc`, `.bash_profile`, `.zshrc`, etc) that modifies the `PATH`?
In .bash_profile:
```sh
export NVM_DIR="/Users/alspohn/.nvm"
[ -s "$NVM_DIR/nvm.sh" ] && . "$NVM_DIR/nvm.sh" # This loads nvm
```
<!-- if this does not apply, please delete this section -->
- If you are having installation issues, or getting "N/A", what does `curl -I --compressed -v https://nodejs.org/dist/` print out?
% Total % Received % Xferd Average Speed Time Time Time Current
Dload Upload Total Spent Left Speed
0 0 0 0 0 0 0 0 --:--:-- --:--:-- --:--:-- 0* Trying 104.20.23.46...
* Connected to nodejs.org (104.20.23.46) port 443 (#0)
* TLS 1.2 connection using TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256
* Server certificate: *.nodejs.org
* Server certificate: COMODO RSA Domain Validation Secure Server CA
* Server certificate: COMODO RSA Certification Authority
* Server certificate: AddTrust External CA Root
> HEAD /dist/ HTTP/1.1
> Host: nodejs.org
> User-Agent: curl/7.43.0
> Accept: */*
> Accept-Encoding: deflate, gzip
>
< HTTP/1.1 200 OK
< Date: Mon, 01 Jan 2018 17:13:41 GMT
< Content-Type: text/html
< Connection: keep-alive
< Set-Cookie: __cfduid=d8eab5992bf4a8489965a9e0269abfdac1514826821; expires=Tue, 01-Jan-19 17:13:41 GMT; path=/; domain=.nodejs.org; HttpOnly
< CF-Cache-Status: EXPIRED
< Vary: Accept-Encoding
< Expires: Mon, 01 Jan 2018 21:13:41 GMT
< Cache-Control: public, max-age=14400
< Server: cloudflare-nginx
< CF-RAY: 3d671c525ce94219-MSP
< Content-Encoding: gzip
<
0 0 0 0 0 0 0 0 --:--:-- --:--:-- --:--:-- 0
* Connection #0 to host nodejs.org left intact
-bash: HTTP/1.1: No such file or directory
<!-- do not delete the following blank line -->
```sh
```
</details> | needs followup,installing nvm | low | Critical |
285,319,419 | rust | What to do about repr(C, simd)? | Currently, `repr(C, simd)` warns about "incompatible representation hints", but leaving off the `repr(C)` triggers the FFI lint when the SIMD type is used in FFI.
Note that `#[repr(C)] #[repr(simd)]` does not warn only due to #47094. Most uses of SIMD in FFI this in the Rust source tree use two separate repr attributes, which is why they didn't trigger this warning so far.
There's two ways to resolve this:
1. Require `repr(C, simd)` for SIMD types used in FFI.
2. Say `repr(simd)` is sufficient for FFI (in principle -- there are other concerns that keep SIMD-FFI feature gated at the moment) and keep the warning about `repr(C, simd)`. It could optionally restricted to some simd types that are known to correspond to C types on the platform in question (e.g., permit `f32x4` but not `f32x3`).
cc @alexcrichton | C-enhancement,A-FFI,T-compiler,A-SIMD,F-simd_ffi,A-repr | medium | Major |
285,384,621 | vscode | when writing a new CompletionItemProvider it is very hard debug as to why the item is not being offered | Whenever I try to write a new CompletionItemProvider, I always have to bang my head against the table many times because there are many rules which hide your item from the completion menu.
I get that these rules are needed, but it is often frustrating for a developer to just see his completion items disappear. Could we get some kind of `debug` mode for CompletionItemProvider so that we can more easily see what is going on with our Completion items?
It would be great to have it something like:
```javascript
class CompletionItemProvider {
static debug = true
provideCompletionItems(document, position: Position, token) {}
}
```
then VSCode would log all reasons why any completion item from this provider was hidden.
Could we have this feature please? | feature-request,suggest | low | Critical |
285,407,580 | vscode | macOS: alt+i always inserts '^' character | - VSCode Version: 1.19.1 & 1.20.0-insider (f9115349ef09cfef2bed669680d10ac58b490dce)
- OS Version: macOS Sierra 10.12.6
Steps to Reproduce:
1. Assign keyboard shortcut "option+i" to a command such as 'Toggle Sidebar Visibility' or 'Go to Definition':
2. Try the shortcut in the editor. Sometimes the shortcut works (correct behavior), other times, the 'ˆ' character is put in the text without the command working (incorrect behavior), and sometimes the command works but the 'ˆ' character is also inserted (incorrect behavior).
Reproduces without extensions: Yes
#22664 & #22894 might be relevant.
---
> 2022.12: Edited to add the upstream link:
**Upstream issue:** https://bugs.chromium.org/p/chromium/issues/detail?id=887519 | upstream,keybindings,macos,upstream-issue-linked,chromium | medium | Critical |
285,463,984 | go | runtime: memmove sometimes faster than memclrNoHeapPointers | Memory allocation using `make([]int, K)` is surprisingly slow compared to `append(nil, ...)`, even though `append` does strictly more work, such as copying.
```
$ cat a_test.go
package main
import "testing"
const K = 1e6
var escape []int
func BenchmarkMake(b *testing.B) {
for i := 0; i < b.N; i++ {
escape = make([]int, K)
}
}
var empty [K]int
func BenchmarkAppend(b *testing.B) {
for i := 0; i < b.N; i++ {
escape = append([]int(nil), empty[:]...)
}
}
$ go version
go version devel +6317adeed7 Tue Jan 2 13:39:20 2018 +0000 linux/amd64
$ go test -bench=. a_test.go
BenchmarkAppend-12 1000 1208800 ns/op
BenchmarkMake-12 1000 1473106 ns/op
```
While reporting this issue, I initially used an older runtime from December 18 in which the effect was much stronger: 10x-20x slowdown. But that seems to have been fixed.
Curiously, this issue is the exact opposite of the problem reported in https://github.com/golang/go/issues/14718 (now closed). | Performance,NeedsInvestigation,compiler/runtime | low | Major |
285,479,083 | flutter | Flutter Inspector on device UI polish: visual change when entering select mode. | Show a mode change when first entering inspect mode (there’s no visual difference currently, so it’s not always clear to the user that clicking the inspect button did anything).
For example, show target icons in the corners of the screen.
| c: new feature,framework,f: inspector,P2,team-framework,triaged-framework | low | Minor |
285,479,239 | flutter | Display of debug paint for selected widgets in the inspector | This will require a refactor of the binding code it flutter so that we have hooks in the appropriate places. User story is visually seeing padding, alignment, and other debug paint data for the selected widget where it is easily understandable as compared to the current behavior of seeing debug paint information for all widgets where the data displayed is overwhelming and often overlapping. | c: new feature,framework,f: inspector,P3,team-framework,triaged-framework | low | Critical |
285,479,569 | flutter | Better debugging of overflow issues using the inspector | Make sure red overflow error areas are always easily clickable bringing you directly to a description of the issue.
Also consider making it possible to jump directly from an overflow exception in the console to inspecting the associated node.
Git clone https://github.com/sergiandreplace/planets-flutter
Try to run it against Android with flutter run
Get this error: A RenderFlex overflowed by 2.0 pixels on the right.
Expect to be able to select the error area on device and jump to the appropriate widget and error property in the inspector.
Example overflow issue YAQS 4933234833227776
| framework,f: inspector,P3,team-framework,triaged-framework | low | Critical |
285,483,865 | youtube-dl | Add support for Adobe Pass Auth TV Provider Midco | ## Please follow the guide below
- [x ] I've **verified** and **I assure** that I'm running youtube-dl **2017.12.31**
### Before submitting an *issue* make sure you have:
- [x ] At least skimmed through the [README](https://github.com/rg3/youtube-dl/blob/master/README.md), **most notably** the [FAQ](https://github.com/rg3/youtube-dl#faq) and [BUGS](https://github.com/rg3/youtube-dl#bugs) sections
- [ x] [Searched](https://github.com/rg3/youtube-dl/search?type=Issues) the bugtracker for similar issues including closed ones
### What is the purpose of your *issue*?
- [ ] Bug report (encountered problems with youtube-dl)
- [ ] Site support request (request for adding support for a new site)
- [ x] Feature request (request for a new functionality)
- [ ] Question
- [ ] Other
---
I noticed that my cable provider (Midcontinent) is not supported by youtube-dl although it appears to be adobe-pass tv provider. It is 49% owned by comcast so I'm hoping it's similar in some way for the setup but not sure. My credentials only work on midco.net and not comcast's xfinity for example. Any help would be appreciated. Thank you so much. :)
| tv-provider-account-needed | low | Critical |
285,484,862 | rust | Incorrect warning about unused const | In the following program:
```
const SIZE: usize = 4;
fn _f(arr: [i32; SIZE]) {
for x in arr.iter() {
println!("{}", x);
}
}
fn main() {}
```
I get a warning about `SIZE` being unused:
```
warning: constant item is never used: `SIZE`
--> src/main.rs:1:1
|
1 | const SIZE: usize = 4;
| ^^^^^^^^^^^^^^^^^^^^^^
|
```
It seems to me that this warning isn't correct, since `SIZE` is used in the signature of `_f` | C-enhancement,A-lints,A-diagnostics,T-compiler,WG-diagnostics | medium | Major |
285,487,141 | pytorch | Assert that some tests must not be skipped under certain CI configurations | It seems pretty easy to accidentally misconfigure one of our CI environments, and consequently fail to run a test. It would be great if we could somehow either:
1. Assert that all tests eventually get run under some configuration (not sure how to do this)
2. Assert that in a given CI environment, some test must not be skipped
That way, an env misconfiguration turns into an error, rather than just some more tests start getting skipped.
cc @ezyang @gchanan @zou3519 @bdhirsh @heitorschueroff @seemethere @malfet @walterddr @pytorch/pytorch-dev-infra @mruberry @VitalyFedyunin | high priority,module: ci,module: tests,triaged,quansight-nack | low | Critical |
285,492,046 | flutter | Build Android artifacts on Windows | Currently, we can only build host artifacts (like `gen_snapshot` and `flutter_tester`) on Windows. We do not support building artifacts on Windows that are supposed to run on other platforms (e.g. the engine for Android). | c: new feature,engine,platform-windows,P3,team-engine,triaged-engine | low | Minor |
285,536,561 | rust | Diagnostics for conversions erroneously say that 'as' only works between primitive types. | See https://play.rust-lang.org/?gist=7194dac428b4355e3f2872e9c0206800&version=stable
The note says "an `as` expression can only be used to convert between primitive types. Consider using the `From` trait", but as show by the expression `Baz as i32`, this is not true. | C-enhancement,A-diagnostics,T-compiler,D-papercut,D-incorrect | low | Minor |
285,567,129 | go | syscall: use Windows FILE_FLAG_BACKUP_SEMANTICS in remaining places (Open) | Hi,
`FILE_FLAG_BACKUP_SEMANTICS` is an optional flag to the Windows `CreateFile` API. It allows accessing files when privilege would be denied by normal Windows ACLs, except for that the user account has the `SE_BACKUP_NAME` privilege. If the calling user account does not have this privilege, adding this flag has no impact to behaviour nor performance.
Go already uses `FILE_FLAG_BACKUP_SEMANTICS` in some CreateFile calls (e.g. `Readlink`, `Utimes`, `UtimesNano`, from issues #8090 and #10804) but not in other cases (notably `Open`). So it's inconsistent.
I would like to request that `FILE_FLAG_BACKUP_SEMANTICS` is added to the `Open` call in `src/syscall/syscall_windows.go`.
As well as fixing the inconsistency, this would allow Go users to open files on disk, in those cases where their permission to do so comes via `SE_BACKUP_NAME` instead of windows ACLs.
Thanks
mappu
More information:
- https://github.com/giuliano108/SeBackupPrivilege
- https://msdn.microsoft.com/en-us/library/windows/desktop/bb530716.aspx
### What version of Go are you using (`go version`)?
`go version go1.9.2 windows/amd64`
### Does this issue reproduce with the latest release?
Yes | OS-Windows,NeedsInvestigation | low | Major |
285,738,440 | flutter | use package: references in the assets section of a pubspec file | You can currently refer to assets from another package using the `packages/package_name/asset_name` format. We should move to using a `package:package_name/asset_name` format - this will be the same as other url references (like imports). | tool,c: API break,a: assets,P2,team-tool,triaged-tool | low | Major |
285,769,824 | rust | Some debuginfo tests are not running | I noticed when adding a debuginfo test that nothing I did caused the test to fail. Tracing back this seems to have been caused by 3e6c83d which broke parsing of the command/check lines, leaving all tests passing without any checking.
PR #47155 runs them again and ignores the failing ones
| A-testsuite,A-debuginfo,T-compiler,C-bug | low | Critical |
285,776,448 | go | cmd/go: pass CGO_FLAGS to C compiler when compiling _cgo_main.c or gccgo | I'm unable to compile `go-sqlite3` with `gccgo` for the PowerPC plataform.
### What version of Go are you using (`go version`)?
`go version go1.8.3 gccgo (Ubuntu 7.2.0-8ubuntu3) 7.2.0 linux/amd64`
### Does this issue reproduce with the latest release?
I have not tested with the latest since the above is the highest version available as pre-compiled package in Ubuntu
### What operating system and processor architecture are you using (`go env`)?
Cross-compiling to ppc with gccgo
```
GOARCH="ppc"
GOBIN=""
GOEXE=""
GOHOSTARCH="amd64"
GOHOSTOS="linux"
GOOS="linux"
GOPATH="/go"
GORACE=""
GOROOT="/usr"
GOTOOLDIR="/usr/lib/gcc/x86_64-linux-gnu/7"
GCCGO="/usr/bin/powerpc-linux-gnuspe-gccgo"
CC="powerpc-fsl-linux-gnuspe-gcc"
GOGCCFLAGS="-m32 -mcpu=8548 -mabi=spe -mspe -mfloat-gprs=double --sysroot=/opt/fsl/2.0/sysroots/ppce500v2-fsl-linux-gnuspe -fPIC -pthread -fmessage-length=0 -fdebug-prefix-map=/tmp/go-build540675113=/tmp/go-build -gno-record-gcc-switches"
CXX="powerpc-fsl-linux-gnuspe-g++"
CGO_ENABLED="1"
PKG_CONFIG="pkg-config"
CGO_CFLAGS="-m32 -mcpu=8548 -mabi=spe -mspe -mfloat-gprs=double --sysroot=/opt/fsl/2.0/sysroots/ppce500v2-fsl-linux-gnuspe -O2 -pipe -g -feliminate-unused-debug-types"
CGO_CPPFLAGS=""
CGO_CXXFLAGS="-g -O2"
CGO_FFLAGS="-g -O2"
CGO_LDFLAGS="--sysroot=/opt/fsl/2.0/sysroots/ppce500v2-fsl-linux-gnuspe -Wl,-O1 -Wl,--hash-style=gnu -Wl,--as-needed"
```
C compiler:
```
powerpc-fsl-linux-gnuspe-gcc --version
powerpc-fsl-linux-gnuspe-gcc (GCC) 4.9.2
```
### What did you do?
Attempt to install go-sqlite3:
`GCCGO=powerpc-linux-gnuspe-gccgo CGO_ENABLED=1 GOARCH=ppc CGO_ENABLED=1 go-7 get -x github.com/mattn/go-sqlite3`
### What did you expect to see?
Package compiling
### What did you see instead?
Compilation fails with
```
In file included from cgo-c-prolog-gccgo:1:0:
/opt/fsl/2.0/sysroots/x86_64-fslsdk-linux/usr/lib/powerpc-fsl-linux/gcc/powerpc-fsl-linux/4.9.2/include/stdint.h:9:26: fatal error: stdint.h: No such file or directory
# include_next <stdint.h>
^
compilation terminated.
```
The compiler I'm using requires some flags to work, in particular the `--sysroot` flag. This is specified in the CC variable which currently is:
```
powerpc-fsl-linux-gnuspe-gcc -m32 -mcpu=8548 -mabi=spe -mspe -mfloat-gprs=double --sysroot=/opt/fsl/2.0/sysroots/ppce500v2-fsl-linux-gnuspe
```
The above works with the older `go version go1.4.2 gccgo (Ubuntu 5.4.0-6ubuntu1~16.04.5) 5.4.0 20160609 linux/amd64`. However, with the version indicated above, the compilation fails because go does not include the flags specified in the CC variable. I managed to work around this by setting CGO_CFLAGS and CGO_LDFLAGS, as can be viewed above. However, I **still** get a compilation error, because apparently gccgo (or cgo?) is not including the CGO_CFLAGS when compiling `_cgo_defun.c` as can be viewed in the verbose output below:
```
GCCGO=powerpc-linux-gnuspe-gccgo CGO_ENABLED=1 GOARCH=ppc CGO_ENABLED=1 go-7 get -x github.com/mattn/go-sqlite3
cd .
git clone https://github.com/mattn/go-sqlite3 /go/src/github.com/mattn/go-sqlite3
cd /go/src/github.com/mattn/go-sqlite3
git submodule update --init --recursive
cd /go/src/github.com/mattn/go-sqlite3
git show-ref
cd /go/src/github.com/mattn/go-sqlite3
git submodule update --init --recursive
cd .
git clone https://go.googlesource.com/net /go/src/golang.org/x/net
cd /go/src/golang.org/x/net
git submodule update --init --recursive
cd /go/src/golang.org/x/net
git show-ref
cd /go/src/golang.org/x/net
git submodule update --init --recursive
WORK=/tmp/go-build601413480
mkdir -p $WORK/golang.org/x/net/context/_obj/
mkdir -p $WORK/golang.org/x/net/
cd /go/src/golang.org/x/net/context
/usr/bin/powerpc-linux-gnuspe-gccgo -I $WORK -c -g -fgo-pkgpath=golang.org/x/net/context -fgo-relative-import-path=_/go/src/golang.org/x/net/context -o $WORK/golang.org/x/net/context/_obj/_go_.o ./context.go ./go17.go ./pre_go19.go
ar rc $WORK/golang.org/x/net/libcontext.a $WORK/golang.org/x/net/context/_obj/_go_.o
mkdir -p /go/pkg/gccgo_linux_ppc/golang.org/x/net/
cp $WORK/golang.org/x/net/libcontext.a /go/pkg/gccgo_linux_ppc/golang.org/x/net/libcontext.a
mkdir -p $WORK/github.com/mattn/go-sqlite3/_obj/
mkdir -p $WORK/github.com/mattn/
cd /go/src/github.com/mattn/go-sqlite3
CGO_LDFLAGS="--sysroot=/opt/fsl/2.0/sysroots/ppce500v2-fsl-linux-gnuspe" "-Wl,-O1" "-Wl,--hash-style=gnu" "-Wl,--as-needed" "-ldl" /usr/lib/gcc/x86_64-linux-gnu/7/cgo -objdir $WORK/github.com/mattn/go-sqlite3/_obj/ -importpath github.com/mattn/go-sqlite3 -gccgo -gccgopkgpath=github.com/mattn/go-sqlite3 -- -I $WORK/github.com/mattn/go-sqlite3/_obj/ -m32 -mcpu=8548 -mabi=spe -mspe -mfloat-gprs=double --sysroot=/opt/fsl/2.0/sysroots/ppce500v2-fsl-linux-gnuspe -O2 -pipe -g -feliminate-unused-debug-types -std=gnu99 -DSQLITE_ENABLE_RTREE -DSQLITE_THREADSAFE=1 -DSQLITE_ENABLE_FTS3 -DSQLITE_ENABLE_FTS3_PARENTHESIS -DSQLITE_ENABLE_FTS4_UNICODE61 -DSQLITE_TRACE_SIZE_LIMIT=15 -DSQLITE_DISABLE_INTRINSIC -Wno-deprecated-declarations -I. backup.go callback.go error.go sqlite3.go sqlite3_context.go sqlite3_load_extension.go sqlite3_other.go sqlite3_type.go
cd $WORK
powerpc-fsl-linux-gnuspe-gcc -fdebug-prefix-map=a=b -c trivial.c
powerpc-fsl-linux-gnuspe-gcc -gno-record-gcc-switches -c trivial.c
cd /go/src/github.com/mattn/go-sqlite3
powerpc-fsl-linux-gnuspe-gcc -I . -fPIC -pthread -fmessage-length=0 -fdebug-prefix-map=$WORK=/tmp/go-build -gno-record-gcc-switches -I $WORK/github.com/mattn/go-sqlite3/_obj/ -m32 -mcpu=8548 -mabi=spe -mspe -mfloat-gprs=double --sysroot=/opt/fsl/2.0/sysroots/ppce500v2-fsl-linux-gnuspe -O2 -pipe -g -feliminate-unused-debug-types -std=gnu99 -DSQLITE_ENABLE_RTREE -DSQLITE_THREADSAFE=1 -DSQLITE_ENABLE_FTS3 -DSQLITE_ENABLE_FTS3_PARENTHESIS -DSQLITE_ENABLE_FTS4_UNICODE61 -DSQLITE_TRACE_SIZE_LIMIT=15 -DSQLITE_DISABLE_INTRINSIC -Wno-deprecated-declarations -I. -o $WORK/github.com/mattn/go-sqlite3/_obj/_cgo_export.o -c $WORK/github.com/mattn/go-sqlite3/_obj/_cgo_export.c
powerpc-fsl-linux-gnuspe-gcc -I . -fPIC -pthread -fmessage-length=0 -fdebug-prefix-map=$WORK=/tmp/go-build -gno-record-gcc-switches -I $WORK/github.com/mattn/go-sqlite3/_obj/ -m32 -mcpu=8548 -mabi=spe -mspe -mfloat-gprs=double --sysroot=/opt/fsl/2.0/sysroots/ppce500v2-fsl-linux-gnuspe -O2 -pipe -g -feliminate-unused-debug-types -std=gnu99 -DSQLITE_ENABLE_RTREE -DSQLITE_THREADSAFE=1 -DSQLITE_ENABLE_FTS3 -DSQLITE_ENABLE_FTS3_PARENTHESIS -DSQLITE_ENABLE_FTS4_UNICODE61 -DSQLITE_TRACE_SIZE_LIMIT=15 -DSQLITE_DISABLE_INTRINSIC -Wno-deprecated-declarations -I. -o $WORK/github.com/mattn/go-sqlite3/_obj/backup.cgo2.o -c $WORK/github.com/mattn/go-sqlite3/_obj/backup.cgo2.c
powerpc-fsl-linux-gnuspe-gcc -I . -fPIC -pthread -fmessage-length=0 -fdebug-prefix-map=$WORK=/tmp/go-build -gno-record-gcc-switches -I $WORK/github.com/mattn/go-sqlite3/_obj/ -m32 -mcpu=8548 -mabi=spe -mspe -mfloat-gprs=double --sysroot=/opt/fsl/2.0/sysroots/ppce500v2-fsl-linux-gnuspe -O2 -pipe -g -feliminate-unused-debug-types -std=gnu99 -DSQLITE_ENABLE_RTREE -DSQLITE_THREADSAFE=1 -DSQLITE_ENABLE_FTS3 -DSQLITE_ENABLE_FTS3_PARENTHESIS -DSQLITE_ENABLE_FTS4_UNICODE61 -DSQLITE_TRACE_SIZE_LIMIT=15 -DSQLITE_DISABLE_INTRINSIC -Wno-deprecated-declarations -I. -o $WORK/github.com/mattn/go-sqlite3/_obj/callback.cgo2.o -c $WORK/github.com/mattn/go-sqlite3/_obj/callback.cgo2.c
powerpc-fsl-linux-gnuspe-gcc -I . -fPIC -pthread -fmessage-length=0 -fdebug-prefix-map=$WORK=/tmp/go-build -gno-record-gcc-switches -I $WORK/github.com/mattn/go-sqlite3/_obj/ -m32 -mcpu=8548 -mabi=spe -mspe -mfloat-gprs=double --sysroot=/opt/fsl/2.0/sysroots/ppce500v2-fsl-linux-gnuspe -O2 -pipe -g -feliminate-unused-debug-types -std=gnu99 -DSQLITE_ENABLE_RTREE -DSQLITE_THREADSAFE=1 -DSQLITE_ENABLE_FTS3 -DSQLITE_ENABLE_FTS3_PARENTHESIS -DSQLITE_ENABLE_FTS4_UNICODE61 -DSQLITE_TRACE_SIZE_LIMIT=15 -DSQLITE_DISABLE_INTRINSIC -Wno-deprecated-declarations -I. -o $WORK/github.com/mattn/go-sqlite3/_obj/error.cgo2.o -c $WORK/github.com/mattn/go-sqlite3/_obj/error.cgo2.c
powerpc-fsl-linux-gnuspe-gcc -I . -fPIC -pthread -fmessage-length=0 -fdebug-prefix-map=$WORK=/tmp/go-build -gno-record-gcc-switches -I $WORK/github.com/mattn/go-sqlite3/_obj/ -m32 -mcpu=8548 -mabi=spe -mspe -mfloat-gprs=double --sysroot=/opt/fsl/2.0/sysroots/ppce500v2-fsl-linux-gnuspe -O2 -pipe -g -feliminate-unused-debug-types -std=gnu99 -DSQLITE_ENABLE_RTREE -DSQLITE_THREADSAFE=1 -DSQLITE_ENABLE_FTS3 -DSQLITE_ENABLE_FTS3_PARENTHESIS -DSQLITE_ENABLE_FTS4_UNICODE61 -DSQLITE_TRACE_SIZE_LIMIT=15 -DSQLITE_DISABLE_INTRINSIC -Wno-deprecated-declarations -I. -o $WORK/github.com/mattn/go-sqlite3/_obj/sqlite3.cgo2.o -c $WORK/github.com/mattn/go-sqlite3/_obj/sqlite3.cgo2.c
powerpc-fsl-linux-gnuspe-gcc -I . -fPIC -pthread -fmessage-length=0 -fdebug-prefix-map=$WORK=/tmp/go-build -gno-record-gcc-switches -I $WORK/github.com/mattn/go-sqlite3/_obj/ -m32 -mcpu=8548 -mabi=spe -mspe -mfloat-gprs=double --sysroot=/opt/fsl/2.0/sysroots/ppce500v2-fsl-linux-gnuspe -O2 -pipe -g -feliminate-unused-debug-types -std=gnu99 -DSQLITE_ENABLE_RTREE -DSQLITE_THREADSAFE=1 -DSQLITE_ENABLE_FTS3 -DSQLITE_ENABLE_FTS3_PARENTHESIS -DSQLITE_ENABLE_FTS4_UNICODE61 -DSQLITE_TRACE_SIZE_LIMIT=15 -DSQLITE_DISABLE_INTRINSIC -Wno-deprecated-declarations -I. -o $WORK/github.com/mattn/go-sqlite3/_obj/sqlite3_context.cgo2.o -c $WORK/github.com/mattn/go-sqlite3/_obj/sqlite3_context.cgo2.c
powerpc-fsl-linux-gnuspe-gcc -I . -fPIC -pthread -fmessage-length=0 -fdebug-prefix-map=$WORK=/tmp/go-build -gno-record-gcc-switches -I $WORK/github.com/mattn/go-sqlite3/_obj/ -m32 -mcpu=8548 -mabi=spe -mspe -mfloat-gprs=double --sysroot=/opt/fsl/2.0/sysroots/ppce500v2-fsl-linux-gnuspe -O2 -pipe -g -feliminate-unused-debug-types -std=gnu99 -DSQLITE_ENABLE_RTREE -DSQLITE_THREADSAFE=1 -DSQLITE_ENABLE_FTS3 -DSQLITE_ENABLE_FTS3_PARENTHESIS -DSQLITE_ENABLE_FTS4_UNICODE61 -DSQLITE_TRACE_SIZE_LIMIT=15 -DSQLITE_DISABLE_INTRINSIC -Wno-deprecated-declarations -I. -o $WORK/github.com/mattn/go-sqlite3/_obj/sqlite3_load_extension.cgo2.o -c $WORK/github.com/mattn/go-sqlite3/_obj/sqlite3_load_extension.cgo2.c
powerpc-fsl-linux-gnuspe-gcc -I . -fPIC -pthread -fmessage-length=0 -fdebug-prefix-map=$WORK=/tmp/go-build -gno-record-gcc-switches -I $WORK/github.com/mattn/go-sqlite3/_obj/ -m32 -mcpu=8548 -mabi=spe -mspe -mfloat-gprs=double --sysroot=/opt/fsl/2.0/sysroots/ppce500v2-fsl-linux-gnuspe -O2 -pipe -g -feliminate-unused-debug-types -std=gnu99 -DSQLITE_ENABLE_RTREE -DSQLITE_THREADSAFE=1 -DSQLITE_ENABLE_FTS3 -DSQLITE_ENABLE_FTS3_PARENTHESIS -DSQLITE_ENABLE_FTS4_UNICODE61 -DSQLITE_TRACE_SIZE_LIMIT=15 -DSQLITE_DISABLE_INTRINSIC -Wno-deprecated-declarations -I. -o $WORK/github.com/mattn/go-sqlite3/_obj/sqlite3_other.cgo2.o -c $WORK/github.com/mattn/go-sqlite3/_obj/sqlite3_other.cgo2.c
powerpc-fsl-linux-gnuspe-gcc -I . -fPIC -pthread -fmessage-length=0 -fdebug-prefix-map=$WORK=/tmp/go-build -gno-record-gcc-switches -I $WORK/github.com/mattn/go-sqlite3/_obj/ -m32 -mcpu=8548 -mabi=spe -mspe -mfloat-gprs=double --sysroot=/opt/fsl/2.0/sysroots/ppce500v2-fsl-linux-gnuspe -O2 -pipe -g -feliminate-unused-debug-types -std=gnu99 -DSQLITE_ENABLE_RTREE -DSQLITE_THREADSAFE=1 -DSQLITE_ENABLE_FTS3 -DSQLITE_ENABLE_FTS3_PARENTHESIS -DSQLITE_ENABLE_FTS4_UNICODE61 -DSQLITE_TRACE_SIZE_LIMIT=15 -DSQLITE_DISABLE_INTRINSIC -Wno-deprecated-declarations -I. -o $WORK/github.com/mattn/go-sqlite3/_obj/sqlite3_type.cgo2.o -c $WORK/github.com/mattn/go-sqlite3/_obj/sqlite3_type.cgo2.c
powerpc-fsl-linux-gnuspe-gcc -I . -fPIC -pthread -fmessage-length=0 -fdebug-prefix-map=$WORK=/tmp/go-build -gno-record-gcc-switches -I $WORK/github.com/mattn/go-sqlite3/_obj/ -m32 -mcpu=8548 -mabi=spe -mspe -mfloat-gprs=double --sysroot=/opt/fsl/2.0/sysroots/ppce500v2-fsl-linux-gnuspe -O2 -pipe -g -feliminate-unused-debug-types -std=gnu99 -DSQLITE_ENABLE_RTREE -DSQLITE_THREADSAFE=1 -DSQLITE_ENABLE_FTS3 -DSQLITE_ENABLE_FTS3_PARENTHESIS -DSQLITE_ENABLE_FTS4_UNICODE61 -DSQLITE_TRACE_SIZE_LIMIT=15 -DSQLITE_DISABLE_INTRINSIC -Wno-deprecated-declarations -I. -o $WORK/github.com/mattn/go-sqlite3/_obj/sqlite3-binding.o -c ./sqlite3-binding.c
powerpc-fsl-linux-gnuspe-gcc -Wall -g -I $WORK/github.com/mattn/go-sqlite3/_obj/ -I /usr/pkg/include -o $WORK/github.com/mattn/go-sqlite3/_obj/_cgo_defun.o -D GOOS_linux -D GOARCH_ppc -D "GOPKGPATH=\"github_com_mattn_go_sqlite3\"" -c $WORK/github.com/mattn/go-sqlite3/_obj/_cgo_defun.c
# github.com/mattn/go-sqlite3
In file included from cgo-c-prolog-gccgo:1:0:
/opt/fsl/2.0/sysroots/x86_64-fslsdk-linux/usr/lib/powerpc-fsl-linux/gcc/powerpc-fsl-linux/4.9.2/include/stdint.h:9:26: fatal error: stdint.h: No such file or directory
# include_next <stdint.h>
^
compilation terminated.
```
Observe that the last compiler invocation did not include the contents of CGO_CFLAGS, preventing `--sysroot` being passed which messes the include path and makes compilation fail.
| NeedsFix | low | Critical |
285,778,810 | rust | Command-line arguments are cloned a lot on Unix | The `std::sys::unix::args` module does a lot of allocation and cloning of command-line parameters:
1. On startup, `std::sys::unix::args::init` copies all of the command-line arguments into a `Box<Vec<Vec<u8>>>` (except on macOS and iOS).
2. When `std::env::args` or `args_os` is called, it eagerly copies all of the args into a new `Vec<OsString>`.
On non-Apple systems, this means there is at least one allocation and clone per argument (plus 2 additional allocations, for the outer `Vec` and `Box`) even if they are never accessed. These extra allocations take up space on the heap for the duration of the program.
On both Apple and non-Apple systems, accessing any args causes at least one *additional* allocation and clone of every arg. Calling `std::env::args` more than once causes *all* arguments to be cloned again, even if the caller doesn't iterate through all of them.
On Windows, for comparison, each arg is cloned lazily only when it is yielded from the iterator, so there are zero allocations or clones for args that are never accessed (update: at least, no clones in Rust code; see comments below).
| I-slow,O-linux,C-enhancement,T-libs | low | Major |
285,832,746 | TypeScript | export as namespace doesn't support nesting namespaces | <!-- BUGS: Please use this template. -->
<!-- QUESTIONS: This is not a general support forum! Ask Qs at http://stackoverflow.com/questions/tagged/typescript -->
<!-- SUGGESTIONS: See https://github.com/Microsoft/TypeScript-wiki/blob/master/Writing-Good-Design-Proposals.md -->
<!-- Please try to reproduce the issue with `typescript@next`. It may have already been fixed. -->
I'm using rollup to generate a UMD module from my TypeScript source. It's part of a bigger product, so one component exports to a root namespace (e.g. `mylib`) and another exports to a namespace nested under that (e.g. `mylib.plugins.myplugin`). Rollup is generating the necessary UMD logic to walk down from the global scope creating namespaces as needed, but I can't model that in my TypeScript `d.ts` file.
The `export as namespace` syntax is working great in the first case, but not the second. It looks like TypeScript doesn't support nested namespaces for this purpose. Is this by design or just an omission?
**TypeScript Version:** 2.7.0-dev.20180103
**Code**: Two files, a `d.ts` containing an `export as namespace foo.bar` declaration and a script that references it.
*declaration.d.ts*
```ts
export var baz;
export as namespace foo.bar;
```
*app.ts*
```ts
console.log(foo.bar.baz);
```
Compile with: `tsc .\test.d.ts .\main.ts`
**Expected behavior:** The file compiles correctly to the following JS:
```js
console.log(foo.bar.baz);
```
**Actual behavior:** Error
```
declaration.d.ts(3,24): error TS1005: ';' expected.
```
If I change `declaration.d.ts` to use `export as namespace foo` (and update `app.ts` as needed), the compilation succeeds.
| Suggestion,In Discussion | low | Critical |
285,842,688 | flutter | Consistent .gitignore for Dart and Flutter projects | We should have a standard and consistent `.gitignore` file common to Dart and Flutter that combines the [recommendations currently on GitHub](https://github.com/github/gitignore/blob/master/Dart.gitignore) with those generated by flutter create (which are different).
From [https://github.com/github/gitignore/blob/master/Dart.gitignore](https://github.com/github/gitignore/blob/master/Dart.gitignore):
```
# See https://www.dartlang.org/tools/private-files.html
# Files and directories created by pub
.packages
.pub/
build/
# If you're building an application, you may want to check-in your pubspec.lock
pubspec.lock
# Directory created by dartdoc
# If you don't generate documentation locally you can remove this line.
doc/api/
```
From `flutter create`:
```
.DS_Store
.atom/
.idea
.packages
.pub/
build/
ios/.generated/
packages
pubspec.lock
.flutter-plugins
```
Other relevant fodder:
https://github.com/github/gitignore/blob/master/Global/JetBrains.gitignore
https://github.com/github/gitignore/blob/master/Global/VisualStudioCode.gitignore
https://github.com/github/gitignore/blob/master/Global/Xcode.gitignore | tool,customer: crowd,P3,team-tool,triaged-tool | high | Critical |
285,909,561 | vscode | [html] extract class name to css file | Steps to Reproduce:
1. click <html> tag;
2. light icon showing, click to show extract menu:
menuitem: extract class name to style tag;
menuitem: extract class name to css file;
demo:
```
<!DOCTYPE html>
<html>
<head>
</head>
<body>
<div class="a">
<div class="a-1">
</div>
</div>
<div class="b">
<div class="b-1">
</div>
</div>
</body>
</html>
```
extract result:
```
.a {
}
.a-1 {
}
.b {
}
.b-1 {
}
``` | feature-request,html | low | Minor |
285,919,749 | rust | Stabilize -Z sanitize | Hi,
I would really like to see -Z sanitize options available in stable rust.
Today they are available in nightly. I have used them for 6 months in the 389 Directory Server project in a production ldap environment as part of the code IO event path. I have exercise the leak and address sanitize options extensively, used them to find many FFI safety issues, and tested multiple build types (exec, dylib, more).
The sanitizer support is very important for FFI assertions during development where a code base uses Rust and C in some form. It allows strict analysis of behaviour, and finding where items are not droped for example.
An important reason for inclusion in stable is *distributions*. As a project, we often have the requirement to build debug builds with the compiler that ships in a distribution to reproduce problems. That means rustc stable that ships in products like fedora or RHEL.
If -Z sanitize isn't in stable, we can't achieve this - and this adds a barrier to adoption. It adds a hurdle to our community inclusiveness as potential developers have more complex steps to follow to create a dev environment. I want to say "dnf install -y cargo rustc ..." and someone should be able to work.
Many people have put a lot of time into sanitizer support, and today it's locked behind nightly. :(
I would love to see this in stable as it brings great run-time analysis features to rust especially for FFI use cases, and having paths to stabilise debugging options into stable rustc will strongly aid adoptions of rust in distributions of linux. It's an advertised feature of the language on email lists, so to have this accessible to anyone would be a huge benefit.
I'm happy to do the coding to enable this, as well as documentation and examples, but I need some help and guidance to achieve this,
Thanks,
See also:
https://github.com/rust-lang/rfcs/issues/670
https://github.com/rust-lang/rust/pull/42711
https://users.rust-lang.org/t/howto-sanitize-your-rust-code/9378
| T-compiler,C-feature-request,A-sanitizers | medium | Critical |
285,976,744 | kubernetes | Modifying nodeSelector on StatefulSet doesn't reschedule Pods | /kind bug
**What happened**:
Changing nodeSelector of a StatefulSet doesn't trigger rescheduling of it's existing pods. I `kubectl apply` the StatefulSet below, and wait for it's Pods to get scheduled onto Nodes with label `node_type: type1`. Then I change the nodeSelector label to `node_type: type2` and do `kubectl apply` again.
```
apiVersion: apps/v1beta1
kind: StatefulSet
metadata:
name: sstest
labels:
app: sstest
spec:
replicas: 2
serviceName: "service"
template:
metadata:
labels:
app: sstest
spec:
nodeSelector:
node_type: type1
containers:
- name: nginx
image: k8s.gcr.io/nginx-slim:0.8
```
**What you expected to happen**:
I expect the Pods to be rescheduled to the type2 Nodes, but nothing happens. The Pods only get rescheduled if I manually kill them using `kubectl delete pod sstest-0`.
**How to reproduce it (as minimally and precisely as possible)**:
1. Create a cluster in which some nodes have the label `node_type: type1` while some other nodes have the label `node_type: type2`.
2. Apply the above StatefulSet definition to the cluster.
3. Change the nodeSelector from `node_type: type1` to `node_type: type2` in the StatefulSet definition file.
4. Apply the file again.
5. Kill a Pod manually to verify that it gets rescheduled to another node.
**Anything else we need to know?**:
I tested the same scenario with Deployments. In that case the rescheduling worked as expected.
**Environment**:
- Kubernetes version (use `kubectl version`):
```
Client Version: version.Info{Major:"1", Minor:"8", GitVersion:"v1.8.4", GitCommit:"9befc2b8928a9426501d3bf62f72849d5cbcd5a3", GitTreeState:"clean", BuildDate:"2017-11-20T05:28:34Z", GoVersion:"go1.8.3", Compiler:"gc", Platform:"darwin/amd64"}
Server Version: version.Info{Major:"1", Minor:"8+", GitVersion:"v1.8.4-gke.1", GitCommit:"04502ae78d522a3d410de3710e1550cfb16dad4a", GitTreeState:"clean", BuildDate:"2017-12-08T17:24:53Z", GoVersion:"go1.8.3b4", Compiler:"gc", Platform:"linux/amd64"}
```
- Cloud provider or hardware configuration: GKE
- OS (e.g. from /etc/os-release): cos | kind/bug,sig/scheduling,sig/apps,lifecycle/frozen | medium | Critical |
286,007,863 | pytorch | [docs] Return value type hint to be specified in docs for torch.is_grad_enabled (i.e. -> bool) | Including master docs: https://pytorch.org/docs/master/search.html?q=is_grad_enabled
cc @jlin27 | module: docs,triaged | low | Major |
286,027,948 | youtube-dl | [bilibili] error 466 in downloading | ### Make sure you are using the *latest* version: run `youtube-dl --version` and ensure your version is *2017.12.31*. If it's not, read [this FAQ entry](https://github.com/rg3/youtube-dl/blob/master/README.md#how-do-i-update-youtube-dl) and update. Issues with outdated version will be rejected.
- [x] I've **verified** and **I assure** that I'm running youtube-dl **2017.12.31**
### What is the purpose of your *issue*?
- [x] Bug report (encountered problems with youtube-dl)
- [ ] Site support request (request for adding support for a new site)
- [ ] Feature request (request for a new functionality)
- [ ] Question
- [ ] Other
---
Hello,i try to use youtube-dl ver12.31 to download a bilibili video but i get a 403 error
youtube-dl -F http://www.bilibili.com/video/av17843872/
[BiliBili] 17843872: Downloading webpage
[BiliBili] 17843872: Downloading video info page
ERROR: Unable to download JSON metadata: HTTP Error 403: Forbidden (caused by HT
TPError()); please report this issue on https://yt-dl.org/bug . Make sure you ar
e using the latest version; type youtube-dl -U to update. Be sure to call yout
ube-dl with the --verbose flag and include its complete output. | bug | medium | Critical |
286,117,271 | neovim | jobstart({string}) fails if 'shell' has unescaped space | Prompted from https://github.com/vim-airline/vim-airline/issues/1595 . If neovim is ran from Git Bash which (by default) exists under `C:\Program Files\Git\usr\bin\bash` functions such as `jobstart` will break due to using the `shell` variable when running the command.
```sh
# We get this
C:\Program Files\Git\usr\bin\bash git
# Instead of
C:\Program\ Files\Git\usr\bin\bash git
```
```vim
:echo &shell
C:\Program Files\Git\usr\bin\bash.exe
```
I don't know if doing the escaping is actually possible, or if there is a better way of solving this but as long spaces in the shell path (which is set internally by nvim) do not break things I'd consider that a nice improvement. | bug,job-control,complexity:low | low | Major |
286,129,107 | rust | Reset rustc's scoped TLS state during unwinding. | This should only happen in the event of an ICE, but right now, if a `Drop` implementation (for e.g. a field in `Session`) would try to access, say, the `TLS_TCX`, it would use-after-free the global `TyCtxt`. | C-cleanup,T-compiler,A-thread-locals | low | Minor |
286,144,548 | flutter | Buttons on iOS do not use squircle paths | @abarth brings this up from time to time (I think mostly jokingly). That CupertinoButton shouldn't actually be a rounded rect, but rather a squircle to match UIKit-based apps. | platform-ios,framework,a: fidelity,f: cupertino,P2,team-design,triaged-design | low | Critical |
286,168,599 | go | runtime: make defs_$GOOS.go regeneratable | It seems that `defs_$GOOS.go` used to generate `defs_$GOOS_$GOARCH.go` in the old days.
https://github.com/golang/go/blob/596e3d9c0176db442a51202a2ae2834ac892d594/src/runtime/defs_darwin.go#L7-L12
However, `-cdefs` is not a valid cgo option today and people started to add features ad hoc.
I think making those regeneratable again have some worthy. | NeedsInvestigation,compiler/runtime | low | Major |
286,221,231 | rust | Option map won't compile while match works well | Consider this snippet:
```rust
fn xxx() -> Option<&'static str> {
Some("word1 word2 word3")
}
fn to_words<'a>() -> Option<Box<Iterator<Item = &'a str> + 'a>> {
xxx().map(|x|Box::new(x.split(' ')))
}
fn main() {
println!("{}", to_words().unwrap().count());
}
```
it won't compiles with the error:
```
error[E0308]: mismatched types
--> src/main.rs:6:5
|
5 | fn to_words<'a>() -> Option<Box<Iterator<Item = &'a str> + 'a>> {
| ------------------------------------------ expected `std::option::Option<std::boxed::Box<std::iter::Iterator<Item=&'a str> + 'a>>` because of return type
6 | xxx().map(|x|Box::new(x.split(' ')))
| ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ expected trait std::iter::Iterator, found struct `std::str::Split`
|
= note: expected type `std::option::Option<std::boxed::Box<std::iter::Iterator<Item=&'a str> + 'a>>`
found type `std::option::Option<std::boxed::Box<std::str::Split<'_, char>>>`
```
but match in exactly same case compiles well:
```rust
fn xxx() -> Option<&'static str> {
Some("word1 word2 word3")
}
fn to_words<'a>() -> Option<Box<Iterator<Item = &'a str> + 'a>> {
match xxx() {
Some(x) => Some(Box::new(x.split(' '))),
None => None
}
}
fn main() {
println!("{}", to_words().unwrap().count());
}
```
And if I explicitly define closure's argument type and returning value type it also compiles well:
```rust
fn xxx() -> Option<&'static str> {
Some("word1 word2 word3")
}
fn to_words<'a>() -> Option<Box<Iterator<Item = &'a str> + 'a>> {
let f: fn(x: &'a str) -> Box<Iterator<Item = &'a str>> = |x| {
Box::new(x.split(' '))
};
xxx().map(f)
}
fn main() {
println!("{}", to_words().unwrap().count());
}
```
I am sure there should not be differences between match and map at all.
This issue is reproduceble on rust stable and nightly.
| C-enhancement,T-lang,A-coercions | low | Critical |
286,291,624 | opencv | Problem compiling both Python2 and 3 support on macOS | <!--
If you have a question rather than reporting a bug please go to http://answers.opencv.org where you get much faster responses.
If you need further assistance please read [How To Contribute](https://github.com/opencv/opencv/wiki/How_to_contribute).
This is a template helping you to create an issue which can be processed as quickly as possible. This is the bug reporting section for the OpenCV library.
-->
##### System information (version)
<!-- Example
- OpenCV => 3.4.0
- Operating System / Platform => macOS Sierra 10.12.6
- Compiler => Apple LLVM version 8.1.0 (clang-802.0.42)
-->
- OpenCV => 3.4.0
- Operating System / Platform => macOS Sierra 10.12.6
- Compiler => Apple LLVM version 8.1.0 (clang-802.0.42)
##### Detailed description
When trying to compile both Python 2 and 3 support compilation fails for Python 3.
I use Macports with latest Python 2.7 and 3.6 installed in /opt/local
Apparently the flags.make file generated in modules/python3/CMakeFiles/opencv_python3.dir does not include in CXX_INCLUDES the directory of Python3_INCLUDE of CMake
Also the same file for Python2 does not include it but since it adds the system directory /usr/include/python2.7 it compiles.
The only PYTHONx_INCLUDE directory that is added to flags.make is that of NUMPY.
I would expect that all the parameters given in the CMAKE GUI where reported in the flags.make file
##### Steps to reproduce
Get core and contrib 3.4.0 source zip files
Set compilation of both Python2 and 3 on macOS 10.12.6
Set /opt/local/Library/Frameworks/Python.framework/Versions/3.6/Headers/ as PYTHON3_INCLUDE_DIR and /opt/local/Library/Frameworks/Python.framework/Versions/2.7/Headers/ as PYTHON2_INCLUDE_DIR
Compile and get a missing Python.h include for cv.so when compiling the Python3 version
Workaround: manually edit build/modules/python3/CMakeFiles/opencv_python3.dir/flags.make to add -I/opt/local/Library/Frameworks/Python.framework/Versions/3.6/Headers/
| category: build/install,incomplete | low | Critical |
286,293,339 | vue | [SSR] Support inline resource for specified files for server side rendering. | ### What problem does this feature solve?
When we are using SSR with CommonsChunkPlugin, we often generate a `manifest.js` at the same time, but `manifest.js` is always very small (about 1kb).
Without SSR using `html-webpack-plugin` and its inline resource plugin, we are able to make `manifest.js` transformed inline into html file.
But for now we can not do for SSR that because the html output is generated by `vue-server-render` automatically.
### What does the proposed API look like?
```js
createBundleRenderer(bundle, {
inlineResources: [] // string or RegExp
})
```
<!-- generated by vue-issues. DO NOT REMOVE --> | feature request | low | Major |
286,313,072 | angular | [Animations] query doesn't work when using the async pipe | ## I'm submitting a...
<!-- Check one of the following options with "x" -->
```
[ x ] Bug report
```
## Current behavior
Animation query doesn't work when using the async pipe.
## Minimal reproduction of the problem with instructions
https://stackblitz.com/edit/query-async
There are 5 sections in the demo. Each container has the trigger on it. The items in all 5 sections should have the same stagger animation applied (including the rows of the MatTable). But only (1) and (4) work.
## Environment
```
Local environment:
Angular CLI: 1.6.3
Node: 8.9.4
OS: win32 x64
Angular: 5.1.3
StackBlitz: [at]angular/animations 5.1.3 (because I added them myself, it's not in the initial scaffolding) and the rest of the packages are 5.0.0, generatd by StackBlitz (I don't know how to change them, but the issue is the same locally with all packages being 5.1.3)
```
| type: bug/fix,area: animations,freq2: medium,P3 | low | Critical |
286,328,101 | rust | Using deprecated items within the crate that deprecated them issues warnings | When deprecating an item, you typically still want the helper methods supporting it to work without the compiler whining that the item is deprecated.
I tried this code: [playground](https://play.rust-lang.org/?gist=de706d2986d8b1cd12a7852e29bc5c1c)
#[deprecated]
pub struct Foo;
impl Foo {
fn foo() -> Foo { Foo }
}
I expected to see this happen: The code compiles without deprecation warnings.
Instead, this happened: Litterally every usage of `Foo` gets a use of deprecated item `Foo` warning.
The fact that I'm deprecating an item doesn't mean the crate itself should now be spammed with deprecation warnings. Working around it by either a global `#![allow(deprecated)]` will swallow legitimate deprecation warnings or require all uses to be tagged with it.
## Meta
Playground, all channels (how do you get the version info from the playground?)
| A-lints,T-lang,C-bug,L-deprecated | low | Major |
286,328,956 | react | Consider removing mouseenter/mouseleave polyfill | As suggested in https://github.com/facebook/react/pull/10247.
Not sure we want to do it, but I decided to create an issue to track future attempts (the PR is stale). | Component: DOM,Type: Breaking Change,React Core Team | low | Major |
286,372,826 | go | cmd/gofmt: quadratic handling of large addition expressions | Given a program with a large string addition
var x = 1 + 1 + 1 + ... 1
gofmt takes time quadratic in the size of the expression to print it back out. Note that this is an integer addition, not a string addition: the problem does not involve string concatenation.
go get -u rsc.io/tmp/bigprog
bigprog gofmt
prints on my system:
1000 int 0.035s strbal 0.008s strbigbal 0.030s str 0.026s strbig 0.047s
2000 int 0.071s strbal 0.012s strbigbal 0.051s str 0.079s strbig 0.115s
4000 int 0.291s strbal 0.022s strbigbal 0.092s str 0.271s strbig 0.362s
8000 int 1.049s strbal 0.032s strbigbal 0.173s str 1.092s strbig 1.249s
16000 int 4.332s strbal 0.070s strbigbal 0.346s str 4.338s strbig 4.852s
32000 int 18.897s strbal 0.112s strbigbal 0.662s str 18.212s strbig 18.889s
64000 int 82.845s strbal 0.224s strbigbal 1.292s str 84.138s strbig 83.718s
128000 int 382.780s strbal 0.425s strbigbal 2.588s str 381.781s strbig 392.788s
See #23222 for full description of this table, but the important part here is that the int, str, and strbig columns are addition chains like above, while the faster strbal and strbigbal columns have parentheses added to make the addition parse trees balanced. The balanced times double nicely as the input size doubles, so there is no problem with actually generating large outputs. In contrast the unbalanced inputs quadruple as input size doubles, a clear quadratic slowdown. The obvious guess is that it is in the code that decides how to format expressions, and probably in the code that decides whether to insert spaces around expressions. (If so, it's my fault and I apologize.)
This is a fairly minor issue, but it would make gofmt 4X faster even on concatenations of size 1000, which are plausible in generated code (I started this after finding one of size 729).
/cc @griesemer | NeedsFix | low | Major |
286,421,328 | pytorch | Bind in Python _backward ATen functions | At the moment they are not bound because we don't support `std::array<bool, N>` binding, but this should be easy to add.
cc @ezyang @SsnL @albanD @zou3519 @gqchen | module: autograd,triaged | low | Minor |
286,422,128 | vscode | [html] code completion replaces text after cursor | - VSCode Version: 1.19.1
- OS Version: MacOS Sierra
Steps to Reproduce:
1. Put some html in and html page with some text `<div>Hello my name is ....</div>`
2. Then before hello type a tag (ie strong "<str" )
3. Hit tab to have it expand strong.
4. Notice that hello is consumed.
Here is a video that shows the bug in action.
https://dev-dynactivesoftware.appspot.com/?blobWSK=ahdzfmRldi1keW5hY3RpdmVzb2Z0d2FyZXKkAgsSEUJsb2JzdG9yZU1ldGFkYXRhIowCQU1JZnY5NVpoMndHX0RzVVJyaFJObFN2OV9pOWlBR2dsQm5BMHhzNWRFT1ZELS04Rmd2SnI2Mk52N0VESFI4S0Z2MnhZLUI1RFp0NGFUT3pOa3BvUnpFRUR0MHJFQy1PQ0JlQkRlalFYVmdZR0poV1liVUJfWDhXTUpjYmRJYXVVckFBTG1BZjhJODM1NFBLVDhuY0NDOWdkb05HN0F5NkZhVm9XT3hUcThVakVUNHZZUERIWTVHcHI3LXZUZjE2Y2Ixay02aVNtQXdYN0EybkdfNmF3aTUyajlxRm9wcDU4eUtodHo3dEVEd0FKdXFsckE5UFVwUGZsX3k3UVE0TTRPZnJKWnZXTHd0RwyiAQZkcy5jbXM
<!-- Launch with `code --disable-extensions` to check. -->
Does this issue occur when all extensions are disabled?: Yes/No
I didn't check because I couldn't figure out how to do this. I opened the command typed the suggested line and I got.
-bash: code: command not found
| feature-request,html | low | Critical |
286,449,936 | kubernetes | Setting defaultMode is not Fully Respected When Pod.spec.securityContext.runAsUser is Set | **Is this a BUG REPORT or FEATURE REQUEST?**:
/kind bug
**What happened**:
Setting `Pod.spec.securityContext.runAsUser` causes the group read permission bit to be set on secrets exposed via volumes, even if `Pod.spec.volumes[x].defaultMode` is set to `256`.
See also: https://github.com/openshift/origin/issues/16424
**What you expected to happen**:
Given a `defaultMode` of `256` the file mode should be `0400` but it is `0440` instead.
**How to reproduce it (as minimally and precisely as possible)**:
Create the following objects and observe the logs of the created pod:
```
---
apiVersion: v1
data:
test: dGVzdA==
kind: Secret
metadata:
name: test-secret
type: Opaque
---
apiVersion: v1
kind: Pod
metadata:
generateName: issue-repro-
spec:
securityContext:
runAsUser: 1000
fsGroup: 1000
containers:
- image: busybox
name: busybox
imagePullPolicy: IfNotPresent
args:
- "ls"
- "-alR"
- "/tmp/dummy-secret"
volumeMounts:
- mountPath: /tmp/dummy-secret
name: test-secret
volumes:
- name: test-secret
secret:
defaultMode: 256
secretName: test-secret
```
**Anything else we need to know?**:
**Environment**:
- Kubernetes version (use `kubectl version`):
```
$ kubectl version
Client Version: version.Info{Major:"1", Minor:"9", GitVersion:"v1.9.0", GitCommit:"925c127ec6b946659ad0fd596fa959be43f0cc05", GitTreeState:"clean", BuildDate:"2017-12-16T03:15:38Z", GoVersion:"go1.9.2", Compiler:"gc", Platform:"darwin/amd64"}
Server Version: version.Info{Major:"1", Minor:"8", GitVersion:"v1.8.0", GitCommit:"0b9efaeb34a2fc51ff8e4d34ad9bc6375459c4a4", GitTreeState:"clean", BuildDate:"2017-11-29T22:43:34Z", GoVersion:"go1.9.1", Compiler:"gc", Platform:"linux/amd64"}
```
- Cloud provider or hardware configuration:
```
$ minikube version
minikube version: v0.24.1
```
- OS (e.g. from /etc/os-release):
```
$ cat /etc/os-release
NAME=Buildroot
VERSION=2017.02
ID=buildroot
VERSION_ID=2017.02
PRETTY_NAME="Buildroot 2017.02"
```
- Kernel (e.g. `uname -a`):
```
$ uname -a
Linux minikube 4.9.13 #1 SMP Thu Oct 19 17:14:00 UTC 2017 x86_64 GNU/Linux
```
- Install tools:
- Others:
| kind/bug,sig/storage,priority/important-longterm,lifecycle/frozen,triage/accepted | high | Critical |
286,475,970 | electron | Add --disable-smooth-scrolling command line switch. | <!--
Thanks for opening an issue! A few things to keep in mind:
- The issue tracker is only for bugs and feature requests.
- Before reporting a bug, please try reproducing your issue against
the latest version of Electron.
- If you need general advice, join our Slack: http://atom-slack.herokuapp.com
-->
* Electron version: 1.7.9
* Operating system: macOS
This is a feature request. I feel that Chromium's smooth scrolling feature is annoying. It would be nice if that could be turned off by some command line switch. Currently, this command line switch is not supported by electron
| enhancement :sparkles:,platform/macOS,platform/linux | medium | Critical |
286,480,453 | godot | Engine.GetSingleton() fails to resolve auto loaded nodes in C# | **Godot version:**
`master` / 0e6e98a65f26c915b5717992ba881cf33406c8bf
**OS/device including version:**
_Manjaro Linux 17.1-rc2_
**Issue description:**
If I have an auto loaded node named `RootContext`, running the following code fails to resolve the node instance with an error, _"Failed to retrieve non-existent singleton 'RootContext'"_:
```c#
var context = Engine.GetSingleton("RootContext");
```
However, the following code correctly resolves and retrieves the node correctly in the same settings (invoked inside another `Node` instance):
```c#
var context = GetNode("/root/RootContext");
```
I'm not entirely sure if it's the correct way to resolve singletons in C# though. | topic:core,documentation | low | Critical |
286,481,430 | TypeScript | Better error for semicolon after decorator | <!-- BUGS: Please use this template. -->
<!-- QUESTIONS: This is not a general support forum! Ask Qs at http://stackoverflow.com/questions/tagged/typescript -->
<!-- SUGGESTIONS: See https://github.com/Microsoft/TypeScript-wiki/blob/master/Writing-Good-Design-Proposals.md -->
**TypeScript Version:** 2.7.0-dev.20171230
**Code**
```typescript
import { component } from "babyioc";
import { Member } from "./Effects";
class Example {
@component(Member);
public readonly member: Member;
}
```
**Expected behavior:**
```
file.ts(6,23): error TSXXXX: Semicolons not allowed after decorators.
```
**Actual behavior:**
```
file.ts(6,23): error TS1146: Declaration expected.
```
Not very important IMO, but a little confusing. | Suggestion,Help Wanted,Domain: Error Messages | low | Critical |
286,508,661 | opencv | Python bindings: objects sharing between multiple threads | OpenCV C++ algorithms/objects are not thread-safe by default due performance and coding simplification reasons.
Concurrent usage of OpenCV objects in multiple threads leads to application crashes.
It is OpenCV usage / programming issue, but it confuses OpenCV users.
Improvement options are:
- integrate automatic locks into generated Python wrappers for OpenCV objects and its methods. Perhaps with timeout runtime parameter and appropriate message on potential "deadlock".
- add detection check with appropriate message if objects is created in one thread, and then it is used from another thread. Implement runtime option to configure this.
- add annotation for Python bindings generator to eliminate such checks for thread-safe OpenCV objects: some define like CV_THREADSAFE or doxygen comment.
**Note**: sharing OpenCV itself between processes via "fork" doesn't work too.
Related issue: https://github.com/opencv/opencv_contrib/issues/1508 ( /cc @Sahloul ) | feature,category: python bindings,RFC | low | Critical |
286,516,296 | rust | Using items from a deprecated module does not warn | Attempting to `use deprecated_module;` or `use deprecated_module::*;` fails, but doing `use deprecated_module::SpecificItem;` compiles with no error. Right now in Diesel I'm actually seeing an even more confusing error:
> use of deprecated item 'types::__test_reexports'
Even though there is nothing in the project with that name. However, I'm unable to reproduce that issue outside of Diesel.
Regardless, the following code compiles without warnings, but should warn.
```rust
#[deprecated(since = "0.1.0")]
pub mod foo {
pub use std::option::Option;
}
pub mod bar {
use foo::Option;
}
```
| A-lints,A-resolve,T-compiler,C-bug,L-deprecated | low | Critical |
286,516,605 | rust | A deprecated module containing tests emits an incredibly confusing deprecation warning that is impossible to silence | This code demonstrates the issue in question:
```rust
#[deprecated(since = "0.1.0")]
mod foo {
#[test]
fn stuff() {}
}
```
Attempting to run `cargo test` on this will give an incredibly opaque error message: `use of deprecated item 'foo::__test_reexports'`. I had to resort to grepping rustc's codebase to figure out where this was coming from.
Additionally, there is *no* way to silence this error, other than allowing use of deprecated items for the entire crate. Attempting to put `#[allow(deprecated)]` either on the test itself, or on the deprecated module has no effect. | A-lints,A-diagnostics,T-compiler,A-libtest,C-bug,L-deprecated | low | Critical |
286,526,291 | rust | libtest Shouldn't Panic On Unexpected Command-line Arg | Right now, if the user passes a command-line argument to a test or benchmark executable that isn't recognized, it will produce an error like the following:
```
thread 'main' panicked at '"Unrecognized option: \'example\'."', /checkout/src/libtest/lib.rs:278:26
note: Run with `RUST_BACKTRACE=1` for a backtrace.
error: bench failed
```
There are two problems with this. First, it would be nice to give a cleaner error message, since this is an expected case. Second and more importantly, the panic halts the `cargo bench` or `cargo test` run without running other executables.
I've attached an example project showing this. The fake benchmark harness in `benches/example.rs` checks for the argument `--example`. However, the `libtest` harness for the tests in the `src` directory is launched first and panics on seeing the same argument. This means that the naive `cargo bench -- --example` doesn't work as one would expect.
[command-line-example.zip](https://github.com/rust-lang/rust/files/1609277/command-line-example.zip)
This is a problem for me because it means I can't add command-line arguments to Criterion.rs benchmarks without an extra hassle for users unless `libtest` also accepts those same arguments.
| T-dev-tools,A-libtest,C-bug | low | Critical |
286,556,382 | youtube-dl | I can't download from qqmusic | Please help me, one month ago i can download from site, but now i can't
sorry for my english
> youtube-dl https://y.qq.com/n/yqq/song/0048jnGb1B9mpX.html
> [qqmusic] 0048jnGb1B9mpX: Download song detail info
> [qqmusic] 0048jnGb1B9mpX: Retrieve vkey
> [qqmusic] 0048jnGb1B9mpX: Checking m4a video format URL
> [qqmusic] 0048jnGb1B9mpX: m4a video format URL is invalid, skipping
> [qqmusic] 0048jnGb1B9mpX: Checking mp3-320 video format URL
> [qqmusic] 0048jnGb1B9mpX: mp3-320 video format URL is invalid, skipping
> [qqmusic] 0048jnGb1B9mpX: Checking mp3-128 video format URL
> [qqmusic] 0048jnGb1B9mpX: mp3-128 video format URL is invalid, skipping
> ERROR: No video formats found; please report this issue on https://yt-dl.org/bug
> . Make sure you are using the latest version; type youtube-dl -U to update. Be sure to call youtube-dl with the --verbose flag and include its complete output.
| geo-restricted,account-needed | low | Critical |
286,578,894 | vscode | Git - add of file with unicode name fails in Git panel | - VSCode Version: Code 1.19.1 (0759f77bb8d86658bc935a10a64f6182c5a1eeba, 2017-12-19T09:46:02.926Z)
- OS Version: Windows_NT ia32 10.0.16299
- Extensions:
Extension|Author (truncated)|Version
---|---|---
vscode-base64|ada|0.0.1
vscode-database|baj|1.2.0
toml|be5|0.0.3
vscode-markdownlint|Dav|0.12.0
createuniqueid|don|0.0.3
githistory|don|0.2.3
EditorConfig|Edi|0.11.1
git-project-manager|fel|1.4.0
mdmath|goe|2.1.0
vscode-wordcount-cjk|hol|1.0.0
Ionide-FAKE|Ion|1.2.3
Ionide-fsharp|Ion|3.15.7
Ionide-Paket|Ion|1.8.1
copy-markdown-as-html|jer|1.0.0
golang-tdd|joa|0.0.9
Go|luk|0.6.71
reflow-markdown|mar|1.3.0
azure-account|ms-|0.2.2
csharp|ms-|1.13.1
vscode-docker|Pet|0.0.23
vscode-icons|rob|7.19.0
go-test-outline|Rom|0.1.4
rust|rus|0.3.2
vscode-autohotkey|sle|0.2.1
vscode-helm|tec|0.3.0
vscode-lldb|vad|0.7.3
vim|vsc|0.10.8
debug|web|0.21.2
---
Steps to Reproduce:
1. Initialize a git repository and open the folder in VSCode.
2. Create a new file named `defaŭlto.html`.
3. From the VSCode Git panel, use the "+" icon to stage the new file.
4. Error reported: `fatal: pathspec 'c:\…\defaulto.html' did not match any files`
Git log reports that it ran the following command: "git add -A -- c:\…\defaŭlto.html"
Running the same command from the git command line in Powershell works without a problem, but also fails from the normal Windows command prompt. Reviewing StackOverflow and other locations, it appears that this is a limitation of the Windows command prompt, but I thought I would document it here since it has a knock-on effect for the Git panel.
<!-- Launch with `code --disable-extensions` to check. -->
Reproduces without extensions: Yes | bug,help wanted,git | medium | Critical |
286,582,826 | You-Dont-Know-JS | Async & Performance: optimize gen-runner in chapter 4 | In the promise-aware gen runner, to prevent the unnecessary extra tick at the beginning, change:
```js
return Promise.resolve().then(
function handleNext(value){ .. }
);
```
...to:
```js
return Promise.resolve(
(function handleNext(value){ .. })()
);
```
| for second edition | low | Major |
286,628,659 | go | testing: update documentation to be clear about when parallel subtests run | ### What version of Go are you using (`go version`)?
`go version go1.9.2 darwin/amd64`
### Does this issue reproduce with the latest release?
Yes
### What operating system and processor architecture are you using (`go env`)?
```
GOARCH="amd64"
GOBIN=""
GOEXE=""
GOHOSTARCH="amd64"
GOHOSTOS="darwin"
GOOS="darwin"
GOPATH="/Users/zwass/dev/go"
GORACE=""
GOROOT="/usr/local/Cellar/go/1.9.2/libexec"
GOTOOLDIR="/usr/local/Cellar/go/1.9.2/libexec/pkg/tool/darwin_amd64"
GCCGO="gccgo"
CC="clang"
GOGCCFLAGS="-fPIC -m64 -pthread -fno-caret-diagnostics -Qunused-arguments -fmessage-length=0 -fdebug-prefix-map=/var/folders/4k/2q3ctn7d5_xb03_zsb3v6lgr0000gn/T/go-build252941481=/tmp/go-build -gno-record-gcc-switches -fno-common"
CXX="clang++"
CGO_ENABLED="1"
CGO_CFLAGS="-g -O2"
CGO_CPPFLAGS=""
CGO_CXXFLAGS="-g -O2"
CGO_FFLAGS="-g -O2"
CGO_LDFLAGS="-g -O2"
PKG_CONFIG="pkg-config"
```
### What did you do?
`go test -v` with the following code:
```
package main
import (
"fmt"
"sync"
"testing"
)
func TestHang(t *testing.T) {
t.Parallel()
var wg sync.WaitGroup
wg.Add(1)
fmt.Println("added wg")
t.Run("", func(t *testing.T) {
fmt.Println("deferred done")
t.Parallel()
fmt.Println("set parallel")
wg.Done()
})
fmt.Println("waiting")
wg.Wait()
fmt.Println("waiting complete")
}
```
### What did you expect to see?
All statements print and test completes
### What did you see instead?
Test hangs after "deferred done" and "waiting" are printed.
This is explained by the following comment in the [blog post introducing subtests](https://blog.golang.org/subtests).
> A parallel test never runs concurrently with a sequential test and its execution is suspended until its calling test function, that of the parent test, has returned.
But I cannot find this mentioned anywhere in the [documentation for the `testing` package](https://godoc.org/testing).
It seems like we at least need to update those docs with the appropriate details. | Documentation,help wanted,NeedsFix | low | Critical |
286,683,871 | rust | optimization remarks broken, names mangled, no file/line addresses (-C remark=all) | ad9a1daa819bbeb8e643a01167b3b69055b88d57 added support for optimization remarks ````RUSTFLAGS="-C remark=all" cargo build --release````
however the output is very hard to use since source locations are broken and names are mangled.
````
note: optimization analysis for inline at <unknown file>:0:0: _ZN11pathfinding12test_prolog517ha0cfb85674d1470fE can be inlined into _ZN11pathfinding4main17hc3ab21b80a078ef2E with cost=-14065 (threshold=275)
note: optimization remark for inline at <unknown file>:0:0: _ZN11pathfinding12test_prolog517ha0cfb85674d1470fE inlined into _ZN11pathfinding4main17hc3ab21b80a078ef2E
note: optimization analysis for inline at <unknown file>:0:0: _ZN11pathfinding12test_prolog617hf4efcc03917e607cE can be inlined into _ZN11pathfinding4main17hc3ab21b80a078ef2E with cost=-14065 (threshold=275)
note: optimization remark for inline at <unknown file>:0:0: _ZN11pathfinding12test_prolog617hf4efcc03917e607cE inlined into _ZN11pathfinding4main17hc3ab21b80a078ef2E
note: optimization analysis for inline at <unknown file>:0:0: _ZN11pathfinding12test_prolog717h9a70d5bf4650de27E can be inlined into _ZN11pathfinding4main17hc3ab21b80a078ef2E with cost=-13915 (threshold=275)
note: optimization remark for inline at <unknown file>:0:0: _ZN11pathfinding12test_prolog717h9a70d5bf4650de27E inlined into _ZN11pathfinding4main17hc3ab21b80a078ef2E
note: optimization analysis for inline at <unknown file>:0:0: _ZN11pathfinding12test_prolog817h899b896fff77b7eaE can be inlined into _ZN11pathfinding4main17hc3ab21b80a078ef2E with cost=-14065 (threshold=275)
note: optimization remark for inline at <unknown file>:0:0: _ZN11pathfinding12test_prolog817h899b896fff77b7eaE inlined into _ZN11pathfinding4main17hc3ab21b80a078ef2E
note: optimization missed for inline at <unknown file>:0:0: _ZN11pathfinding20print_shortest_paths17hac2d4d48645262b5E will not be inlined into _ZN11pathfinding4main17hc3ab21b80a078ef2E
note: optimization missed for inline at <unknown file>:0:0: _ZN4core3ptr13drop_in_place17h9afff73c9d313742E will not be inlined into _ZN11pathfinding4main17hc3ab21b80a078ef2E
note: optimization floating-point for loop-vectorize at <unknown file>:0:0: loop not vectorized: cannot prove it is safe to reorder floating-point operations
````
cargo 0.25.0-nightly (a88fbace4 2017-12-29)
rustc 1.25.0-nightly (ee220daca 2018-01-07)
| A-LLVM,T-compiler,C-bug | low | Critical |
286,699,110 | pytorch | descriptor 'add' of 'torch._C._VariableBase' object needs an argument | the pytorch document said that the add opeartion can be like this
## ` torch.add(input, value=1, other, out=None)`
but it will rise the error of descriptor 'add' of 'torch._C._VariableBase' object needs an argument
when i do
## `torch.add(input=a, value=0.01, other=a)`
a is a Variable like
Variable containing:
5.0500
[torch.FloatTensor of size 1]
but torch.add(a, 0.01, a) is fine to work
cc @jlin27 | module: docs,triaged | low | Critical |
286,737,886 | rust | Rust's stdio should not ignore EBADF error on non-windows platforms | I've discovered, that `std::io::stdin/out/err()` streams unconditionally ignore `EBADF`-like IO errors on all platforms. This is done by checking the read/write error in a `handle_ebadf()` function.
https://github.com/rust-lang/rust/blob/1ccb50eaa670f86b69e7a64484a8c97e13169183/src/libstd/io/stdio.rs#L123-L128
It appears, that this behavior was first introduced here https://github.com/rust-lang/rust/commit/a7bbd7da4eb98127104fdfd415ad6c746f7e2a12
The commit clearly has Windows in mind, where it appears the standard streams may be unavailable. But on Linux, the streams are expected to be always present, so there's no reason to ignore `EBADF` in the first place, as it indicates that something is very wrong.
Not only that, but due to file descriptor reuse behavior on Unixes, if descriptors 0/1/2 are not open, sometimes the very next calls to `open()` will allocate them. This means, that a program running without properly preallocated 0/1/2 descriptors may start happily `println!()`-ing over its own sqlite database, or send private execution logs across a tcp connection.
So, if `std::io::stdout/err()` happens to discover that something yanked the descriptors from under program's feet, the proper response is not to silently ignore `EBADF`, but to **panic()**, before something else unwittingly allocated it with likely disastrous consequences. | C-enhancement,T-libs-api,A-io | low | Critical |
286,767,880 | angular | [Animations] weird initialisation states | <!--
PLEASE HELP US PROCESS GITHUB ISSUES FASTER BY PROVIDING THE FOLLOWING INFORMATION.
ISSUES MISSING IMPORTANT INFORMATION MAY BE CLOSED WITHOUT INVESTIGATION.
-->
## I'm submitting a...
<!-- Check one of the following options with "x" -->
<pre><code>
[ ] Regression (a behavior that used to work and stopped working in a new release)
[X] Bug report <!-- Please search GitHub for a similar issue or PR before submitting -->
[ ] Feature request
[ ] Documentation issue or request
[ ] Support request => Please do not submit support request here, instead see https://github.com/angular/angular/blob/master/CONTRIBUTING.md#question
</code></pre>
## Current behavior
<!-- Describe how the issue manifests. -->
The Angular's animation system does not initialize the same component (in the exemple 'test-btn') the same way if it is in a transcluded component (in the exemple 'test-popup').
## Expected behavior
<!-- Describe what the desired behavior would be. -->
The transcluded test-btn should display only one text/state.
## Minimal reproduction of the problem with instructions
<!--
For bug reports please provide the *STEPS TO REPRODUCE* and if possible a *MINIMAL DEMO* of the problem via
https://plnkr.co or similar (you can use this template as a starting point: http://plnkr.co/edit/tpl:AvJOMERrnz94ekVua0u5).
-->
http://plnkr.co/edit/nnFiPLz0YyLjfc8Ll2HZ?p=preview
You can click on the "Send" button, it works properly. Toggle the popup, you will see the same button component but all possible states are visible. Note: the aria-hidden props are properly set.
Bonus: if you spam the click, the text come from bottom and not from top as expected via
```
transition('invisible => visible', [
style({opacity: 0, transform: 'translateY(-100%)'}),
animate('0.5s')
])
```
## Environment
<pre><code>
Angular version: 5.1.3 latest
<!-- Check whether this is still an issue in the most recent Angular version -->
</code></pre>
| type: bug/fix,area: animations,freq3: high,P3 | low | Critical |
286,850,094 | go | cmd/pprof: use $GOPATH for source file lookup | ### What version of Go are you using (`go version`)?
go version go1.9.2 darwin/amd64
### Does this issue reproduce with the latest release?
yes
### What operating system and processor architecture are you using (`go env`)?
```
GOARCH="amd64"
GOBIN="/Users/phemmer/.go/bin"
GOEXE=""
GOHOSTARCH="amd64"
GOHOSTOS="darwin"
GOOS="darwin"
GOPATH="/Users/phemmer/.go:/tmp/go"
GORACE=""
GOROOT="/usr/local/go"
GOTOOLDIR="/usr/local/go/pkg/tool/darwin_amd64"
GCCGO="gccgo"
CC="clang"
GOGCCFLAGS="-fPIC -m64 -pthread -fno-caret-diagnostics -Qunused-arguments -fmessage-length=0 -fdebug-prefix-map=/var/folders/xk/jhvsfjg14zqfgtwvm0kpsy61p_63g5/T/go-build517300090=/tmp/go-build -gno-record-gcc-switches -fno-common"
CXX="clang++"
CGO_ENABLED="1"
CGO_CFLAGS="-g -O2"
CGO_CPPFLAGS=""
CGO_CXXFLAGS="-g -O2"
CGO_FFLAGS="-g -O2"
CGO_LDFLAGS="-g -O2"
PKG_CONFIG="pkg-config"
```
### What did you do?
Attempted to profile a heap dump from a binary built on another host and use `list foo`
### What did you expect to see?
The source code with corresponding data for each line.
### What did you see instead?
```
(pprof) list unmarshalBinary
Total: 1.95GB
ROUTINE ======================== github.com/influxdata/kapacitor/vendor/github.com/influxdata/influxdb/models.(*point).unmarshalBinary in /root/go/src/github.com/influxdata/kapacitor/vendor/github.com/influxdata/influxdb/models/points.go
1.49GB 1.49GB (flat, cum) 76.51% of Total
Error: open /root/go/src/github.com/influxdata/kapacitor/vendor/github.com/influxdata/influxdb/models/points.go: no such file or directory
```
On my local system, this source code can instead be found in `/Users/phemmer/.go/...`.
Ref #13231 which is related, but I can't use $GOROOT, as I don't have a full go toolkit in my home directory, only application source. | NeedsInvestigation,compiler/runtime | low | Critical |
286,852,534 | go | cmd/compile: inefficient CALL setup when more than 32bytes of args | ```
$ gotip version
go version devel +a62071a209 Sat Jan 6 04:52:00 2018 +0000 linux/amd64
```
```
type T struct {
s1, s2 string
}
//go:noinline
func foo(t T) { _ = t }
func bar() {
var t T
foo(t)
}
```
generates
```
0x0020 00032 (test.go:14) MOVUPS X0, (SP)
0x0024 00036 (test.go:14) MOVUPS X0, 16(SP)
0x0029 00041 (test.go:14) CALL "".foo(SB)
```
but when
```
type T struct {
s1, s2, s3 string // one more string
}
```
```
0x001d 00029 (test.go:13) XORPS X0, X0
0x0020 00032 (test.go:13) MOVUPS X0, "".t+48(SP)
0x0025 00037 (test.go:13) MOVUPS X0, "".t+64(SP)
0x002a 00042 (test.go:13) MOVUPS X0, "".t+80(SP)
0x002f 00047 (test.go:13) MOVQ SP, DI
0x0032 00050 (test.go:14) LEAQ "".t+48(SP), SI
0x0037 00055 (test.go:14) DUFFCOPY $854
0x004a 00074 (test.go:14) CALL "".foo(SB)
```
The stack is bigger; first we `MOVUPS` a bunch of zeros to `48/64/80(SP)`, then we call `DUFFCOPY` to move them again to (SP). This seems wasteful. Even if we cross the multiple-MOVs/DUFF threshold, it seems it would be possible to just `DUFFZERO` at `(SP)`, essentially the thing the first snippet does.
This also happen when there's no zeroing going on. For example, for `struct { a, b, c, d int64}`, when initialized as `t = {1, 2, 3, 4}`, the values are moved directly to `(SP)`, but for `struct { a, b, c, d, e int64}`, which is bigger than 32bytes, they aren't. There are 5 moves high into the stack and then a `DUFFCOPY` call moves them to `(SP)`.
| Performance,binary-size,compiler/runtime | low | Minor |
286,865,098 | vscode | Add an optional configurable toolbar below the menu | Whole my life I used ide where there was a customizable toolbar. I starded to use vscode and stopped to use it after one day. It is impossible to remember all shorcuts. In any other popular ide you can put any menu item on toolbar and use it from time to time.
**visual vtudio:**
**idea:**
**eclipse:**
**netbeans:**
**code blocks:**
**komodo ide:**
**atom:**
**notepad++:**
**gedit:**
**github:**
WTF MS? VS code? VS users?
Don't tag it "out-of-scope" as you did before with similar requests. Even simple editors have toolbar.
It isn't advanced feature it's basic feature for most people.
Using ide without toolbar is not user friendly for most users. It is vim way.
**People really need it like "exit from vim":**
<br>
<br>
<br>
<br>
<br>
**FAQ FOR PEOPLE WHO ARE AGAINST THE TOOLBAR**:
>I don't want to use toolbar so I am against that feature-request.
It is not a problem "it is optional toolbar". If you don't want you can don't use it.
>If you want it so why don't you form a team and submit a PR?
People have offered, go look at the many other bug reports made on the same issue. MS has said they would likely reject it. They don't even want to add hooks to make a plugin possible. So it's not a resource issue, it's an ideological one.
**Temporary workaround by @GorvGoyl:**
>It seems like it won't be the high priority feature in foreseeable future so I made this extension which adds handy buttons like beautify, list files, undo, redo, save all etc to the editor menu bar in the VSCode. [Shortcut Menu Bar](https://marketplace.visualstudio.com/items?itemName=jerrygoyal.shortcut-menu-bar)
 | feature-request,layout | high | Critical |
286,898,764 | go | dist: provide one-line installer | We all agree we want a 1-line installer. tools/cmd/getgo is a start but has work to do. This is the tracking bug for shipping it (and advertising it on the install page).
| NeedsFix,FeatureRequest,DevExp | medium | Critical |
286,899,301 | godot | UndoRedo error with consecutive create_action calls | <!-- Please search existing issues for potential duplicates before filing yours:
https://github.com/godotengine/godot/issues?q=is%3Aissue
-->
**Godot version:**
99da466
**OS/device including version:**
Ubuntu 17.10
**Issue description:**
After this, I'm done torturing UndoRedo, I think. :)
There doesn't appear to be any way out of a create_action() once you've started it. There's a commit but no rollback. If one expects another create_action() to discard the previous commands, they're treated to an error and a now-broken commit.
**Steps to reproduce:**
```
todd@todd-ab28d3:~/repos/remote/brainsick-godot-tests$ cat tests/qa/classes/UndoRedoTest-uncommit.gd
extends MainLoop
var testobject1
var undoredo
func _initialize():
testobject1 = TestClass.new()
testobject1.set_prop(1)
undoredo = UndoRedo.new()
# main test body
func _iteration(delta):
test_create_uncommit()
test_create_commit()
return true # exit MainLoop
func _finalize():
undoredo.free()
func test_create_uncommit():
undoredo.create_action('first action - method')
undoredo.add_do_method(testobject1, 'set_prop', 2)
undoredo.add_undo_method(testobject1, 'set_prop', testobject1.get_prop())
#undoredo.commit_action()
# where's rollback?
assert(testobject1.prop == 1)
func test_create_commit():
undoredo.create_action('second action - property')
#undoredo.create_action('second action - property', UndoRedo.MERGE_DISABLE)
#undoredo.create_action('second action - property', UndoRedo.MERGE_ENDS)
#undoredo.create_action('second action - property', UndoRedo.MERGE_ALL)
undoredo.add_do_property(testobject1, 'prop', 3)
undoredo.add_undo_property(testobject1, 'prop', testobject1.prop)
undoredo.commit_action()
assert(testobject1.prop == 3)
class TestClass:
var prop
func get_prop():
return prop
func set_prop(prop):
self.prop = prop
todd@todd-ab28d3:~/repos/remote/brainsick-godot-tests$ ../brainsick-godot/bin/godot.x11.tools.64 -s tests/qa/classes/UndoRedoTest-uncommit.gd
No touch devices found
OpenGL ES 3.0 Renderer: AMD Radeon (TM) R9 Fury Series (AMD FIJI / DRM 3.18.0 / 4.13.0-21-generic, LLVM 5.0.0)
GLES3: max ubo light: 409
GLES3: max ubo reflections: 455, ubo size: 144
ARVR: Registered interface: Native mobile
SCRIPT ERROR: test_create_commit: Assertion failed.
At: res://tests/qa/classes/UndoRedoTest-uncommit.gd:40.
ERROR: clear_history: Condition ' action_level > 0 ' is true.
At: core/undo_redo.cpp:326.
```
I have an updated, more robust test at godotengine/godot-tests#3. | topic:core,confirmed,documentation | low | Critical |
286,900,597 | pytorch | Carefully audit contiguity requirements of code | While I was working on the cuDNN bindings, I was a bit nervous about when we required things to be contiguous or not (for the most part, I tried to exactly mimic what the API did previously, but I did not feel very confident about it.) The reason I was nervous is that we don't seem to carefully test our algorithms on all combinations of contiguity for each of their inputs.
Now that we most recently discovered and fixed #4500, it might be worth planning to do an in depth contiguity audit of PyTorch's codebase + test coverage, to see if there are other corners we have missed.
I am vaguely reminded of #3170, which was a analogous bug with sparse coalesced.
cc @csarofeen @ptrblck | module: cudnn,triaged | low | Critical |
286,918,277 | opencv | ORB descriptor cannot use SIFT keypoints | ##### System information (version)
- OpenCV => 3.4.0-dev
- Operating System / Platform => ALL
- Compiler => ALL
##### Detailed description
Although ORB works with FAST or AKAZE keypoints, it fails to use the SIFT ones!
##### Steps to reproduce
```.py
import cv2
sift = cv2.xfeatures2d.SIFT_create()
orb = cv2.ORB_create()
rgb = cv2.imread('any_image.png')
kps = sift.detect(rgb)
kps, descrs = orb.compute(rgb, kps)
```
##### Actual result
Memory allocation error, and a crash! | category: features2d,RFC | low | Critical |
286,949,508 | vscode | [xml] add on enter rules | - VSCode Version: Code 1.19.1 (0759f77bb8d86658bc935a10a64f6182c5a1eeba, 2017-12-19T09:46:23.884Z)
- OS Version: Windows_NT x64 10.0.16299
- Extensions:
Extension|Author (truncated)|Version
---|---|---
xml|Dot|1.9.2
auto-close-tag|for|0.5.5
auto-complete-tag|for|0.0.2
auto-rename-tag|for|0.0.15
beautify|Hoo|1.1.1
csharp|ms-|1.13.1
qub-xml-vscode|qub|1.2.8
---
Steps to Reproduce:
1. Automatic indentation does not work when writing XML(appending child).
2. Automatic indentation works when writing HTML.
<!-- Launch with `code --disable-extensions` to check. -->
Reproduces without extensions: Yes/No | feature-request,languages-basic | low | Minor |
286,958,283 | pytorch | Met 'cudnnDestroyDropoutDescriptor' while run multiply gpu-based models in multiply processes | I loaded a model and replicated 3 models which are all on different GPUs, and use torch.multiprocessing to create 4 processes. By this setting, I run one model in on process, so that I parallel run them. The running is fine, but when the processes were finished, this error occurred:
```
Exception ignored in: <bound method CuDNNHandle.__del__ of <torch.backends.cudnn.CuDNNHandle object at 0x7f86ea1bcf28>>
Traceback (most recent call last):
File "/home/lgy/installed/anacoda3/lib/python3.6/site-packages/torch/backends/cudnn/__init__.py", line 114, in __del__
AttributeError: 'NoneType' object has no attribute 'cudnnDestroy'
```
The code where the error happens is:
```
class DropoutDescriptor(object):
def __del__(self):
check_error(lib.cudnnDestroyDropoutDescriptor(self))
```
My pytorch version:
```
$> pip list
....
torch (0.3.0.post4)
torchtext (0.1.1)
torchvision (0.2.0)
....
```
Is there any ways to fix this?
cc @csarofeen @ptrblck | module: multi-gpu,module: cudnn,triaged | low | Critical |
287,029,058 | flutter | [google_sign_in] Remove guava dependency | This is no bugreport, more a question.
I started editing the google sign in plugin for android and figured out several problem, I would like to fix.
In order to do so, I would like to know what the getToken method is used for. According to the rest of the implementation, you can get the token out of the GoogleSignInAccount after a successful login. IMHO this method is not necessary, but I guess there's a reason for you to have it in there.
If not, I would like to remove it, since it's not really required and the dependency required (guava) is a [method limit](https://developer.android.com/studio/build/multidex.html#about) killer, since guava comes [with about 15k methods](http://www.methodscount.com/?lib=com.google.guava%3Aguava%3A20.0)
If it is needed for some reason, I created an alternative implementation without the guava dependency using android build in tools. See the PR mentioned below
| c: new feature,team,platform-android,p: google_sign_in,package,c: proposal,P2,team-android,triaged-android | low | Critical |
287,041,491 | opencv | `KeyPoint` and `vector<KeyPoint>` are not sufficient for detector/descriptor decoupling | Sorry for the long report in advance, I hope developers can react to this, and agree to find a better approach.
The existence of `Feature2D::detect` and `Feature2D::compute` advertises for detector/descriptor decoupling, in which detected keypoints by virtually any detector shall be describable using any descriptor. Unfortunately, the underlying implementation gives developers a huge challenge to achieve that. Perhaps a slight modification in the interfaces can improve the developer experience.
Currently, `Keypoint` class provides `pt`, `size`, `angle`, `response`, `octave`, and `class_id` fields, but they are insufficient, and many questions remains unanswered.
- What is the scale factor for a keypoint at the octave level, given its diameter (`size`)?
- What is the scale ratio between different octave levels? How to compute/recompute it? Is it a constant ratio, or does it have a specific form?
- Are negative octaves allowed? What does negative values mean?
- What does huge octave values mean? How to relate that to the scale?
- How to interpret the `response`? Is it unified along all detectors?
- Mixing keypoints from different detectors robustifies the match, but how to solve the ambiguities (e.g. octave scale) then? Can we know which keypoint belongs to what?
- Can detectors pass some callback to solve such ambiguities?
While many descriptors tackles these issues to some extent, some other descriptors (e.g. AKAZE) just reject any foreign keypoints. More importantly, each descriptor strives by its own, at different levels of maturity, to convert the keypoints to a format it expects (e.g. ORB as in #10556). It is a pity that, for example, SIFT encodes `layer` and `octave` in the same integer using bitwise operators, instead of having separate fields for each (#4554). Such lack of concrete definition for the keypoints attributes and allowed ranges waste developers time and efforts.
Currently, `Keypoint` class is just a container, and so the `vector` that aggregates the keypoint instances!
I wonder why C++ iterable classes are not utilized, rather than just a vector!? Such iterators are capable of provide all the missing information as well ...
| RFC | low | Major |
287,071,466 | create-react-app | Proposal: explicit named imports for non-JS/CSS assets | ## Problem
We currently allow you to do this:
```js
import logo from './logo.png';
```
After getting used to it, you’ll probably be comfortable with this giving you a URL.
But what about other types? For example, what should this return?
```js
import doc from './doc.md';
```
Markdown source? Compiled HTML? An AST?
What about this?
```js
import Icon from './icon.svg';
```
Should this give you a link? The SVG content? A React component?
The usual answer is “decide it for yourself in the configuration file”. However, that doesn’t work for CRA so we decided to treat all unknown extensions as URL imports. This is not ideal because in some cases it just doesn’t make any sense, and in others there are advanced (but still relatively common) use cases that aren’t satisfied.
## Proposal
What if we allowed to user to pick what they want, from a limited supported subset per filetype?
```js
import { url as logoUrl } from './logo.png';
import { html as docHtml } from './doc.md';
import { ReactComponent as Icon } from './icon.svg';
```
Named imports are checked by webpack so you’d get a compile error if you use an unsupported one.
Things that are unused will be tree shaken so if you only use e.g. HTML of Markdown files, their source won’t be bundled. Same for SVGs (whether you consume them as raw source, URLs, or React components).
Other zero-configuration tools can also adopt this approach.
## Concerns
* What do we do with the default import? Ideally I’d like to forbid it for anything other than JS/CSS because the intent is not clear for asset files (which version do you get?) We could do this with a lint rule.
* If we make the breaking change, how do we update the consumers? We could write a codemod (it should be very simple).
* It would be nice to coordinate this across at least a few other projects (e.g. @ndelangen Storybook). Maybe @KyleAMathews (Gatsby) @devongovett (Parcel) @rauchg (Next) would also be interested? I imagine we’ll need to write a multi-file Webpack loader for this, but I don’t see why other bundlers couldn’t adopt a similar convention.
* Build performance: generating all possible content variations [can be too slow](https://mobile.twitter.com/wSokra/status/950713344163446785). Ideally it would be nice if loaders had information about which import was used.
Thoughts?
| issue: proposal | high | Critical |
287,073,778 | go | cmd/internal/obj/x86: FSAVE assembled as FNSAVE | ### What version of Go are you using (`go version`)?
`go version go1.9.2 linux/amd64`
### Does this issue reproduce with the latest release?
Yes.
### What operating system and processor architecture are you using (`go env`)?
GOARCH="amd64"
GOBIN=""
GOEXE=""
GOHOSTARCH="amd64"
GOHOSTOS="linux"
GOOS="linux"
GOPATH="/home/quasilyte/CODE/intel/avxgen"
GORACE=""
GOROOT="/usr/local/go"
GOTOOLDIR="/usr/local/go/pkg/tool/linux_amd64"
GCCGO="gccgo"
CC="gcc"
GOGCCFLAGS="-fPIC -m64 -pthread -fmessage-length=0 -fdebug-prefix-map=/tmp/go-build350631872=/tmp/go-build -gno-record-gcc-switches"
CXX="g++"
CGO_ENABLED="1"
CGO_CFLAGS="-g -O2"
CGO_CPPFLAGS=""
CGO_CXXFLAGS="-g -O2"
CGO_FFLAGS="-g -O2"
CGO_LDFLAGS="-g -O2"
PKG_CONFIG="pkg-config"
### What did you do?
1. Assemble this file by `go tool asm fsave.s`:
```asm
TEXT main·fsave(SB),0,$0
FSAVE (AX)
RET
```
2. Check output by `go tool objdump fsave.o`
```
TEXT main.fsave(SB) gofile../home/quasilyte/CODE/asm/fsave.s
fsave.s:2 0x94 dd30 FNSAVE 0(AX)
fsave.s:3 0x96 c3 RET
```
3. Note that disassembler prints "FNSAVE" (it's correct). You may also check it with external disassemblers like binutils objdump and Intel XED.
```bash
$ ./obj/examples/xed -64 -d dd30
DD30
ICLASS: FNSAVE CATEGORY: X87_ALU EXTENSION: X87 IFORM: FNSAVE_MEMmem108 ISA_SET: X87
SHORT: fnsave ptr [rax]
```
### What did you expect to see?
FSAVE instruction is assembled into FSAVE.
Result encoding is "9bdd30".
### What did you see instead?
FSAVE instruction is assembled into FNSAVE.
Result encoding is "dd30".
## Additional notes
- It's not only FSAVE/FNSAVE, there are more such instructions. I can provide more-or-less complete list later, potentially in a separate issue.
- It's impossible to select datasize as there is no FNSAVEL/FNSAVES variants, only FNSAVE (which is FSAVE right now). x86csv makes me believe that Go asm should have either suffixed version to make datasize selection possible.
```
$ cat src/x86/x86.csv | egrep 'fnsave|fsave'
"FNSAVE m94/108byte","FNSAVES/FNSAVEL m94/108byte","fnsaves/fnsavel m94/108byte","DD /6","V","V","","","w","",""
"FSAVE m94/108byte","FSAVE m94/108byte","fsave m94/108byte","9B DD /6","V","V","","pseudo","w","",""
```
I can send a CL that fixes this.
The only trouble is backwards-compatibility: if FSAVE becomes FNSAVE, what to do with existing code?
Note that FSAVE is really a combination of FWAIT and FNSAVE.
| NeedsFix | low | Critical |
287,096,512 | godot | Object is not rendered behind Dof Near Blur distance on Android | <!-- Please search existing issues for potential duplicates before filing yours:
https://github.com/godotengine/godot/issues?q=is%3Aissue
-->
**Godot version:**
<!-- Specify commit hash if non-official. -->
4b414f45c
**OS/device including version:**
<!-- Specify GPU model and drivers if graphics-related. -->
Android 7.0 / SM-955N (Galaxy S8+)
**Issue description:**
<!-- What happened, and what was expected. -->
Expected
What happened
Object is not rendered behind Dof Near Blur distance on Android
Dof Far Blur works as expected. but not Dof Near Blur.
**Steps to reproduce:**
1. set Dof Near Blur distance in WorldEnvironment
2. run on Android
**Minimal reproduction project:**
<!-- Recommended as it greatly speeds up debugging. Drag and drop a zip archive to upload it. -->
[Test DOF.zip](https://github.com/godotengine/godot/files/1615263/Test.DOF.zip)
| bug,platform:android,topic:rendering,confirmed,topic:3d | low | Critical |
287,120,986 | vscode | Heterogeneous DPI support on Linux | - VSCode Version: Code 1.19.1 (0759f77bb8d86658bc935a10a64f6182c5a1eeba, 2017-12-19T09:41:01.414Z)
- OS Version: Linux x64 4.14.11-300.fc27.x86_64
Steps to Reproduce:
1. A Wayland desktop session on a multi-display setup with heterogeneous DPI scaling. Mine is a HiDPI laptop display scaled at 200% and an external monitor using 100% scaling, resulting in xwayland scaled at 100% on all displays. gnome-shell is the compositor.
2. Launch Visual Studio Code.
3. Move the IDE window between the displays.
Like any other X11 application, Visual Studio Code cannot adapt to heterogeneous scaling.
Wayland support, discussed in #1739, is needed to enable this. I feel that opening a new issue is justified by the problem described, which was not mentioned by the commenters on that issue.
<!-- Launch with `code --disable-extensions` to check. -->
Does this issue occur when all extensions are disabled?: Yes | feature-request,upstream,linux,electron,workbench-os-integration,upstream-issue-linked | high | Critical |
287,213,298 | pytorch | Mixed Tensor/TensorList arguments in ATen functions with explicit derivatives | Today, you cannot write a derivative for a function with signature `f(Tensor, TensorList)` (or any combination thereof) because our code generation does not support it.
The primary technical barrier to implementing this is the fact that autograd internals works with a completely flat list of tensors. To support mixed Tensor/TensorList arguments in general, one would also have to record enough metadata to reconstruct the intended groupings. This is only required when there are multiple TensorList arguments; with a single TensorList argument the correct grouping can be unambiguously inferred (though it takes some work to do so.)
We should decide what we want to do long term. Is it:
1. Never support mixed Tensor/TensorList; instead, require users to somehow work around the restriction in some way. (For example, `index(Tensor, TensorList)` can be decomposed into a series of differentiable operations that don't require mixed arguments.)
2. Support mixed Tensor plus a single TensorList. In this regime no special metadata is needed, and the wrapping code can look solely at the calling convention to determine how to unwrap
3. Support arbitrary mixed Tensor and TensorList. In this regime some extra metadata is needed to resolve the wrapping/unwrapping. If written in a robust way, this regime could be generalized to support arbitrary nested structures.
An ancillary consideration is ONNX, which also adopts the "single, flat list of tensors" model. ONNX has effectively bet that operators with complicated input tensor structures are unlikely, and you pay a big complexity cost if this is not the case.
cc @ezyang @SsnL @albanD @zou3519 @gqchen | module: autograd,triaged,enhancement | low | Minor |
287,286,835 | go | cmd/compile: conversion error is a bit inaccurate | Please answer these questions before submitting your issue. Thanks!
#### What did you do?
If possible, provide a recipe for reproducing the error.
A complete runnable program is good.
A link on play.golang.org is best.
```
package main
func main() {
s := "s"
_ = 'r' != s
_ = 1 != s
_ = 1.1 != s
}
```
#### What did you expect to see?
I expected to see "untyped rune", "untyped int", "untyped float".
#### What did you see instead?
"type untyped number" which is not a searchable keyword in the https://golang.org/ref/spec.
```
# command-line-arguments
./k.go:5:10: cannot convert 'r' (type untyped number) to type string
./k.go:5:10: invalid operation: 'r' != s (mismatched types rune and string)
./k.go:6:8: cannot convert 1 (type untyped number) to type string
./k.go:6:8: invalid operation: 1 != s (mismatched types int and string)
./k.go:7:10: cannot convert 1.1 (type untyped number) to type string
./k.go:7:10: invalid operation: 1.1 != s (mismatched types float64 and string)
```
To be precise, conversion error here is already inaccurate.
The error message implies `s := string('r')` is incorrect code, but it's not true.
Also "type untyped" is redundant expression to me.
```
cannot compare 'r' (untyped rune) to string
```
or
```
cannot assign 'r' (untyped rune) to string
```
is correct one. (initial one is better in this context)
#### Does this issue reproduce with the latest release (go1.9.2)?
I can see the conversion error, but I can't see "untyped number" in the error message.
#### System details
```
go version devel +23aefcd9ae Wed Jan 10 00:00:40 2018 +0000 darwin/amd64
GOARCH="amd64"
GOBIN=""
GOCACHE="/Users/hiro/Library/Caches/go-build"
GOEXE=""
GOHOSTARCH="amd64"
GOHOSTOS="darwin"
GOOS="darwin"
GOPATH="/Users/hiro/.go"
GORACE=""
GOROOT="/Users/hiro/go"
GOTMPDIR=""
GOTOOLDIR="/Users/hiro/go/pkg/tool/darwin_amd64"
GCCGO="gccgo"
CC="clang"
CXX="clang++"
CGO_ENABLED="1"
CGO_CFLAGS="-g -O2"
CGO_CPPFLAGS=""
CGO_CXXFLAGS="-g -O2"
CGO_FFLAGS="-g -O2"
CGO_LDFLAGS="-g -O2"
PKG_CONFIG="pkg-config"
GOGCCFLAGS="-fPIC -m64 -pthread -fno-caret-diagnostics -Qunused-arguments -fmessage-length=0 -fdebug-prefix-map=/var/folders/wq/dwn8hs0x7njbzty9f68y61700000gn/T/go-build157306342=/tmp/go-build -gno-record-gcc-switches -fno-common"
GOROOT/bin/go version: go version devel +23aefcd9ae Wed Jan 10 00:00:40 2018 +0000 darwin/amd64
GOROOT/bin/go tool compile -V: compile version devel +23aefcd9ae Wed Jan 10 00:00:40 2018 +0000
uname -v: Darwin Kernel Version 17.3.0: Thu Nov 9 18:09:22 PST 2017; root:xnu-4570.31.3~1/RELEASE_X86_64
ProductName: Mac OS X
ProductVersion: 10.13.2
BuildVersion: 17C88
lldb --version: lldb-900.0.64
Swift-4.0
gdb --version: GNU gdb (GDB) 8.0.1
```
| compiler/runtime | low | Critical |
Subsets and Splits
No saved queries yet
Save your SQL queries to embed, download, and access them later. Queries will appear here once saved.