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 |
|---|---|---|---|---|---|---|
101,122,958 | rust | DST coercions fall over with UFCS | ``` rust
#![feature(core, rc_weak)]
use std::cell::RefCell;
use std::rc::{Rc, Weak};
trait Baz { fn get(&self) -> i32; }
impl Baz for i32 { fn get(&self) -> i32 { *self } }
fn main() {
let a: Rc<RefCell<i32>> = Rc::new(RefCell::new(42));
let c1: Weak<RefCell<Baz>> = Rc::downgrade(&a); // Doesn't work
let c2: Weak<RefCell<Baz>> = Rc::downgrade(&a) as Weak<_>; // works
let c3: Weak<RefCell<Baz>> = a.downgrade(); // works
}
```
CC @nrc
| A-DSTs,T-compiler,C-bug | low | Minor |
101,236,314 | java-design-patterns | Data-Context-Interaction (DCI) pattern | **Description:**
The Data-Context-Interaction (DCI) design pattern focuses on separating data storage, business logic, and interaction context within a system. It aims to improve code readability and maintainability by clearly defining these three components:
1. **Data**: Represents the structure and storage of data, typically using simple data objects.
2. **Context**: Defines the environment or situation in which the data operates, establishing roles and interactions for the objects involved.
3. **Interaction**: Contains the business logic and processes that manipulate the data within a specific context.
By implementing the DCI pattern, we aim to enhance the separation of concerns, making the system more modular and easier to understand.
**References:**
- [Project Contribution Guidelines](https://github.com/iluwatar/java-design-patterns/wiki)
- [DCI Architecture Overview](https://en.wikipedia.org/wiki/Data,_Context_and_Interaction)
- [The DCI Architecture: A New Vision of Object-Oriented Programming](http://www.artima.com/articles/dci_vision.html)
- [Video - Re-thinking foundations of OO programming](http://oredev.org/videos/dci--re-thinking-the-foundations-of-oo)
- [DCI in Ruby](http://dci-in-ruby.info/)
**Acceptance Criteria:**
1. Create a new pattern named `data-context-interaction` under the project.
2. Implement classes representing the Data, Context, and Interaction components of the DCI pattern.
3. Provide example usage and tests demonstrating the functionality and benefits of the DCI pattern in the project.
| info: help wanted,epic: pattern,type: feature | low | Major |
101,317,013 | TypeScript | Re-exporting namespace declarations in ES6 ambient declaration | Right now many declaration files take this form:
``` typescript
declare namespace MyLib {
}
declare module 'myLib' {
export = MyLib;
}
```
However, there doesn't seem to be an equivalent for ES6 modules:
``` typescript
declare module 'myLib' {
export default MyLib; // this works for exporting the default, but other exported items in MyLib are not considered to be named exports of the ES6 module
export * from MyLib; // this is essentially what I'm trying to accomplish
}
```
| Suggestion,In Discussion | medium | Major |
101,335,951 | go | go/printer: "Line of stars" detection false positive for commented out block of code with each line dereferencing a pointer. | #### What version of Go are you using (go version)?
```
go version go1.4.2 darwin/amd64
```
#### What operating system and processor architecture are you using?
OS X 10.10.5, amd64.
#### What did you do?
Consider the following program:
``` Go
package main
import "fmt"
func main() {
var i [3]int
p0 := &i[0]
p1 := &i[1]
p2 := &i[2]
*p0 = 5
*p1 = 6
*p2 = 7
fmt.Println("Hello, playground:", *p0, *p1, *p2)
}
```
If you comment out the first block of code and run `gofmt` on it, you get:
``` Go
package main
import "fmt"
func main() {
/*
var i [3]int
p0 := &i[0]
p1 := &i[1]
p2 := &i[2]
*/
*p0 = 5
*p1 = 6
*p2 = 7
fmt.Println("Hello, playground:", *p0, *p1, *p2)
}
```
However, if you comment out the second block of code like this:
``` Go
package main
import "fmt"
func main() {
var i [3]int
p0 := &i[0]
p1 := &i[1]
p2 := &i[2]
/*
*p0 = 5
*p1 = 6
*p2 = 7
*/
fmt.Println("Hello, playground:", *p0, *p1, *p2)
}
```
#### What did you expect to see?
After formatting the [above code](http://play.golang.org/p/XYoKp-oEhk) with `gofmt`, I would expect it to be formatted similarly to when any other piece of code is commented out, so it should be:
``` Go
package main
import "fmt"
func main() {
var i [3]int
p0 := &i[0]
p1 := &i[1]
p2 := &i[2]
/*
*p0 = 5
*p1 = 6
*p2 = 7
*/
fmt.Println("Hello, playground:", *p0, *p1, *p2)
}
```
#### What did you see instead?
``` Go
package main
import "fmt"
func main() {
var i [3]int
p0 := &i[0]
p1 := &i[1]
p2 := &i[2]
/*
*p0 = 5
*p1 = 6
*p2 = 7
*/
fmt.Println("Hello, playground:", *p0, *p1, *p2)
}
```
Which is not consistent with anything.
I'm pretty sure I know why it happens. It has to do with go/printer detecting a so called "line of stars" style of comments, which normally look like this:
``` Go
/*
* Line
* of
* stars.
*/
```
See relevant code at https://github.com/golang/go/blob/467a2cb73ca867ccdb1dd22be220fa51b6d52491/src/go/printer/printer.go#L529-L539.
However, in this case, it's not a comment block because there is no space after the \* as in normally formatted comments, it's just commented out code that happens to falsely trigger the line of stars comment block handling.
| NeedsInvestigation | low | Minor |
101,519,463 | youtube-dl | Add support for universalmusic.fr site | Hi,
Is this possible to add support for this website: http://www.universalmusic.fr/ ?
Here is an example of a video to support: http://www.universalmusic.fr/artiste/10732-calogero/videos/18522-tien-an-men
Thanks :)
| site-support-request | low | Minor |
101,614,766 | youtube-dl | extractor for edx.org courses | Course page: https://courses.edx.org/courses/HarvardX/CS50x3/2015/info
Course wage (list of pages with lectures): https://courses.edx.org/courses/HarvardX/CS50x3/2015/courseware/b2f7d86728354866a2c4438e76c3ec55/
Lecture (actual video): https://courses.edx.org/courses/HarvardX/CS50x3/2015/courseware/b2f7d86728354866a2c4438e76c3ec55/99a3e898bd0c43aab9757d71b114529f/
| site-support-request,account-needed | low | Minor |
101,616,608 | go | cmd/dist: race detector should run test/ tests | There are lots of interesting code patterns in test/ dir. It would be useful to at least build them with race detector in race.bash.
| RaceDetector | low | Minor |
101,690,238 | go | x/mobile: gomobile test | This may turn into a design for a `gomobile test` subcommand, but for now I'm writing down some notes on what I've been thinking about this topic, as I have a tendency to forget parts of this story when I switch work.
Current and potential testing systems for app programmers:
**All-Go apps:**
- `go test` runs unit tests on the host. Works today. Fast, portable, and friendly to development.
- `gomobile test` same UI as `go test` but they run on device. Can this be made easy to configure? (Provisioning issues on iOS, app permission issues on Android, or they would run via adb shell which rules out GUI tests.) Is this useful?
- GUI tests. Right now x/mobile/app/app_test.go has some primitive GUI tests driven by the host. Can these be run on-device as part of `gomobile test`?
**Go libraries used by android apps:**
- `go test` runs unit tests on the host.
- `gomobile test` runs tests on the device. Same discussion as above.
- Integration tests. As Go is the library, these have to be driven by the primary app environment, Android Studio, Xcode, etc. (For example, x/mobile/bind tests do this with gradle from the Android SDK.)
If we created a `gomobile test` mode, it would be built out of the go_darwin_arm_exec and go_android_arm_exec scripts used to run unit tests from the standard library. These are tricky to setup (in particular the ios-deploy dependency) and very tricky to debug. It i not clear we could make that easier with the gomobile tool.
So it’s not clear yet that it is worth building `gomobile test`, but a worthwhile first experiment would be rewriting x/mobile/app/app_test.go to be a unit test running under an .apk. If that is possible, the second step would be to rewrite go_android_arm_exec so that it builds app packages like go_darwin_arm_exec. In some ways it is easier to install and start an app on Android than it is on iOS, but extracting stdio usefully on a device without root (most development devices) is a bit of work.
| mobile | low | Critical |
101,699,440 | go | x/review/git-codereview: does not work with 'git worktree' | See https://github.com/blog/2042-git-2-5-including-multiple-worktrees-and-triangular-workflows for info about 'git worktree'.
At the least, the hook checker needs updating:
``` bash
$ git sync
git-codereview: checking hook: stat /Volumes/Data/src/go.ssa/.git/hooks/commit-msg: not a directory
```
There might be other stuff as well.
| NeedsInvestigation | low | Minor |
101,713,260 | rust | FreeBSD x86_64 test failures | After the new snapshots landed, test failures showed up.
| O-x86_64,O-freebsd,C-bug | low | Critical |
101,839,785 | go | x/tools/go/ssa: expected "Output:" text not instantiated for testing.InternalExample struct | go version go1.5rc1 darwin/amd64
For bonzo_test.go
```
package bonzo
import "fmt"
import _ "testing"
func ExampleHello() {
fmt.Println("hello")
// Output: hello
}
```
x/tools/go/ssa generates code:
```
function: test$main.main
block 0:
t0 = new [1]testing.InternalExample (test main)
t1 = &t0[0:int]
t2 = &t1.Name [#0]
*t2 = "ExampleHello":string
t3 = &t1.F [#1]
*t3 = github.com/tardisgo/tardisgo-samples/errors/exampletests/bonzo.ExampleHello
t4 = slice t0[:]
t5 = testing.Main(main$1, nil:[]testing.InternalTest, nil:[]testing.InternalBenchmark, t4)
return
```
A program demonstrating this SSA code generation is at https://github.com/tardisgo/tardisgo-samples/tree/master/errors/exampletests
The Output field of structure testing.InternalExample is not being set to the text following "// Output:", which means that when the code runs (not currently possible in x/tools/go/ssa/interp) the output is:
```
Input: github.com/tardisgo/tardisgo-samples/errors/exampletests/bonzo
--- FAIL: ExampleHello (0.00s)
got:
hello
want:
FAIL
FAIL
The following tests failed:
github.com/tardisgo/tardisgo-samples/errors/exampletests/bonzo
```
Rather than the test passing as expected:
```
PASS
ok github.com/tardisgo/tardisgo-samples/errors/exampletests/bonzo 0.006s
```
| Tools | low | Critical |
101,856,761 | opencv | Organize project management with github issue | Now that we have migrated from chilliproject to github it could be helpful to evaluate one of this tool at http://liftux.com/posts/using-github-issues-project-management/.
Most of this solutions are free for open source projects (like github account itself). Waffle it is also open source. Not reviewed and opensource also Taiga.io it is interesting at https://taiga.io/
| RFC | low | Minor |
101,872,692 | opencv | videoio last frame forever when usb camera disconnected | Hi I am using opencv 3.0.0 compiled under mingw 4.9.2 and I found out a bug, when the camera is disconnected no empty frame is returned but instead the last captured image from the camera, over and over again. There is no way how to find out if the camera was disconnected even .opened still returns true over the camera.
| bug,priority: normal,category: videoio,affected: 3.4 | low | Critical |
101,946,396 | go | image/jpeg: specify APP1 segment for outputting EXIF data in jpeg.Encode()? | Currently there is no way to output EXIF in jpeg.Encode() so if a JPEG file is modified use image/jpeg the EXIF info is lost.
Reading EXIF using jpeg.Decode() has a similar problem but reopening the file and reading a few kilo bytes is probably an acceptable workaround. Reopening a file to insert a segment is much more cumbersome and inefficient.
One way to do this is to introduce a new field in jpeg.Options. Something like below diff. Thoughts?
```
diff --git a/src/image/jpeg/reader.go b/src/image/jpeg/reader.go
index adf97ab..5942609 100644
--- a/src/image/jpeg/reader.go
+++ b/src/image/jpeg/reader.go
@@ -64,6 +64,7 @@ const (
// but in practice, their use is described at
// http://www.sno.phy.queensu.ca/~phil/exiftool/TagNames/JPEG.html
app0Marker = 0xe0
+ app1Marker = 0xe1
app14Marker = 0xee
app15Marker = 0xef
)
diff --git a/src/image/jpeg/writer.go b/src/image/jpeg/writer.go
index 91bbde3..e404efa 100644
--- a/src/image/jpeg/writer.go
+++ b/src/image/jpeg/writer.go
@@ -541,6 +541,7 @@ const DefaultQuality = 75
// Quality ranges from 1 to 100 inclusive, higher is better.
type Options struct {
Quality int
+ App1Data []byte
}
// Encode writes the Image m to w in JPEG 4:2:0 baseline format with the given
@@ -597,6 +598,12 @@ func Encode(w io.Writer, m image.Image, o *Options) error {
e.buf[0] = 0xff
e.buf[1] = 0xd8
e.write(e.buf[:2])
+
+ // Write APP1 data if specified
+ if o != nil && o.App1Data != nil {
+ e.writeMarkerHeader(app1Marker, 2 + len(o.App1Data))
+ e.write(o.App1Data)
+ }
```
| Thinking | low | Critical |
102,034,903 | TypeScript | Provide suggested fixes for unresolved entities | Something like the following:
See https://github.com/Microsoft/TypeScript-Sublime-Plugin/issues/327 for the original suggestion.
| Suggestion,Help Wanted | low | Minor |
102,035,879 | go | cmd/go,cmd/cgo: link-time symbol resolution for buildmode=c-shared | version:
```
$ go version
go version go1.5 linux/amd64
```
I'm working with a software that loads plugins structured as follows. Plugins implement an interface, and the main program provides an API, so a typical call stack might look like:
```
main_program_function()
\_ plugin_function()
\_ main_program_other_function()
```
Up to now, the main program and the plugins are written in C. I was eager to see if I could start building plugins in Go instead of C, now that 1.5 is out, so I gave it a try with something that looks like:
``` go
package main
// #include "main_program_api.h"
import "C"
//export plugin_function
func plugin_function() {
C.main_program_other_function()
}
func main () {
// unused
}
```
Unfortunately, this fails with:
```
./plugin.go:8: undefined reference to `main_program_other_function'
```
Ian Lance Taylor suggested I pass '--unresolved-symbols' to the linker as a work around. I verified that that works as expected, both when passed via the command line:
```
CGO_LDFLAGS="-Wl,--unresolved-symbols=ignore-all" go build -buildmode=c-shared
```
or via the code
``` go
// #cgo LDFLAGS: -Wl,--unresolved-symbols=ignore-all
```
| NeedsInvestigation,GoCommand,compiler/runtime | low | Major |
102,085,607 | kubernetes | Deprecate Kubelet Read-only port | There are some voices that it's no longer necessary/won't be necessary for original purpose, as Heapster is/will be using secured port to scrape cAdvisor data. @roberthbailey @vishh @bgrant0607.
On the other hand with @fgrzadkowski we used it to debug some problems Filip was having, so it isn't completely useless.
@dchen1107 @yujuhong - any thoughts?
| priority/backlog,area/security,kind/cleanup,sig/node,priority/important-longterm,lifecycle/frozen | medium | Critical |
102,106,082 | opencv | Python bindings naming convention and packaging | I've been working on creating automated builds (using [conda](http://conda.pydata.org/docs/)) of [OpenCV2](https://github.com/menpo/conda-opencv) and [OpenCV3](https://github.com/menpo/conda-opencv3). In particular, I am only really interested in the Python bindings. However, I have a couple of questions about the design decisions for OpenCV's Python bindings:
### `cv2.so` is just dropped inside the Python `site-packages` folder
I don't think that this is good practise. It goes against standard Python packaging principles and dirties the contents of the `site-packages` folder. However, I also have a concrete complaint: on Windows, due to the way shared library loading is achieved, we need to drop all the `opencv*.dll` next to the `cv2.so` module. However, if another poorly packaged project also requires being dropped in the root of `site-packages`, and binds to opencv using C-extensions, it will also need these DLLs. As I mentioned, I'm trying to make automated builds, and this becomes very difficult if every project needs to start dumping files into the same folder, some of which may overwrite each other. Not to mention the fact that removing a package may involve cleaning up dependencies, which would include the aforementioned DLLs (breaking other packages that might rely on them). Therefore, **why not place the `cv2.so` inside a small Python package that just exposes all the same top level modules/methods?** This would have the advantage of isolating all the dependencies and also allow easy shipping of companion pure Python code with OpenCV, if this was ever needed.
### OpenCV2 and OpenCV3 both use `cv2` as their namespace
This one seems particularly crazy to me. Since OpenCV3 deprecated the `cv2.cv` module, which some people require, OpenCV3 is not a drop in replacement for OpenCV2. Which is fine, as long as the two are namespaced separately... but they are not. **Why was OpenCV3 not given a new package name, such as `cv3`?** At the moment, if someone installs OpenCV3 over OpenCV2, they will lose their OpenCV2! This seems very strange to me. If the whole point of the major version change was to denote breaking, non-backwards compatible changes, why would you keep the same namespace? This is further aggravated by the problem mentioned above, in that the `cv2.so` will literally get overwritten if OpenCV3 is installed after OpenCV2 and vice-versa.
As I said, I mostly just wanted to discuss the rationale behind these points. I understand that I do not represent a core contributor, but OpenCV is an amazing project and a massive boon to the Python community. Therefore, I would love to help steer it in a direction that made it more accessible to even more Python developers by providing easy to install automated builds.
| priority: normal,feature,category: build/install,affected: 3.4,RFC | low | Major |
102,136,672 | opencv | Not determenistic behaviour of flann on android smartphone | I'm using the HierarchicalClusteringIndex from the flann module for image recognition. The image features are computed with ORB.
On my Linux desktop computer (Ubuntu 14.04) everything works fine. On my Android smartphone (Galaxy Nexus, Android 4.3) the knnSearch via "findNeighbors" yields to different results after each restart of the application. I tested on a Galaxy S3 in addition, the problem stays the same.
This case occurs both with OpenCV 2.4.X and OpenCV 3.0.
Here are parts of the code to reproduce:
``` cpp
// init flann index
cvflann::Matrix<unsigned char> m_features(words.data, words.rows, words.cols);
Ptr<cvflann::HierarchicalClusteringIndex<cvflann::Hamming<unsigned char> > > kdIndex =
new cvflann::HierarchicalClusteringIndex<cvflann::Hamming<unsigned char> >
(m_features,cvflann::HierarchicalClusteringIndexParams(10));
kdIndex->buildIndex();
// prepare knn-queries
vector<int> indices(1);
vector<int> dists(1);
cvflann::KNNResultSet<int> m_indices(1);
m_indices.init(indices.data(),dists.data());
for(int i = 0; i < query.rows; ++i){
Mat row = query.row(i);
// perform knn-query
kdIndex->findNeighbors(m_indices, row.ptr<unsigned char>(0), cvflann::SearchParams(8));
}
```
where words is an Mat-object with 100 000 rows and 32 columns and type CV_8U. It is used as visual vocabulary. query is Mat-object that consists of image descriptors computed with ORB.
| bug,priority: normal,affected: 2.4,affected: 3.4,category: flann | low | Minor |
102,171,139 | go | runtime: revisit non-constant assist ratio | Currently we compute a conservative assist ratio that assumes the entire heap is reachable in order to prevent heap overshoot caused by underestimating. Unfortunately, this means we usually over-assist (by a factor of ~2 in steady state). This causes the GC cycle to finish below the heap goal and with high CPU utilization. This is okay to do for a few cycles, but the unfortunate part is that it's a stable state for the trigger controller, so we keep doing this (it scales the heap growth by the overutilization factor and finds that it would have been right on target with 25% utilization, so it doesn't change the trigger).
One solution we talked about early in the pacer development, but never put together a concrete design for, was to adjust the assist ratio as we approached the heap goal. This means the mutator may assist more toward the end of the cycle, but it will let us target the heap goal more precisely (sort of a "mid-flight correction"). We need to consider how to do this without allowing overshoot and without putting too much assist strain on the end of the cycle (which increases mutator latency). We also need to consider whether this affects how we compute the trigger error, which currently assumes linearity.
/cc @RLH
| compiler/runtime | low | Critical |
102,197,109 | neovim | PCRE regex support | I'm not asking for https://github.com/neovim/neovim/issues/304 , or even the "refactoring regex matching to a swappable plugin" suggestion within there. As much as I would love that, it seems to me that would either be a headache to plugin writers who'll need to check which style user has and duplicate their regexes, or end up splitting plugins into those usable with vim-regex and those that use PCRE.
My suggestion is to allow PCRE style regular expressions entirely optionally and distinctly from the existing vim-regex matching, so that mere mortals who do not grok vim-re can more easily write and edit regex matches in vim. It need not interfere with current syntax at all, and perhaps could be just functions like `pcre_match(str, pat)` (equivalent to `str =~ pat` except for using PCRE syntax in pat), and `pcre_substitute(str, pat, sub, flags)` (equivalent to `substitute`, with PCRE syntax in pat).
Currently, this can sort-of be achieved with external plugins like https://github.com/othree/eregex.vim using its `E2v` function that converts PCRE-regexes into vim-style regexes, but
(1) it is an external dependency
(2) there are load order issues - only those plugins that load after eregex.vim can use its functions, and crucially, users can't use it in vimrc itself unless you do something hacky like prefix every statement with `au VimEnter`.
(3) perhaps most important of all, using `E2v` repeatedly means taking a potentially enormous performance hit, since every regex has to be first parsed character by character by eregex and then transformed into vim-regex textually, after which the actual matching begins.
So, having a PCRE engine within neovim directly and being able to use it by explicit choice would be a big usability improvement for a lot of us.
| enhancement | high | Major |
102,289,576 | youtube-dl | Get a list of all playlist and descriptions from a channel | Hello,
I am trying to create a podcast from a channel.
Is there a way to get a list of all playlists from a channels (e.g. in the same way we get a list of videos from a playlist) ?
And continuing on that, is there a way to get the playlist description?
-rsd
| request | low | Minor |
102,395,648 | go | cmd/link: combining dwarf on darwin can be very slow | I'm noticing ~15x increase in build time when switched to go 1.5.
It is somehow connected with a package github.com/ry/v8worker - binding for v8 javascript engine.
I'm building on OS X.
This is a trivial program:
``` go
package main
import _ "github.com/ry/v8worker"
func main() {
}
```
**1.4** build time:
real 0m1.075s
user 0m0.964s
sys 0m0.108s
**1.5** build time:
real 0m16.816s
user 0m16.225s
sys 0m1.226s
Go 1.5 spends almost all of the time on the **link** step:
/usr/local/Cellar/go/1.5/libexec/pkg/tool/darwin_amd64/link -o $WORK/slow_build/_obj/exe/a.out -L $WORK -L /Users/ianic/work/web/src/golang/pkg/darwin_amd64 -extld=clang++ -buildmode=exe -buildid=14ca37ec220e553c7bf9829fe97751df5a041f5f $WORK/slow_build.a
**1.4 output for: time go build -v -x**
WORK=/var/folders/dx/p9b300dd27dcbqcrz0h9kn800000gn/T/go-build829145301
slow_build
mkdir -p $WORK/slow_build/_obj/
mkdir -p $WORK/slow_build/_obj/exe/
cd /Users/ianic/work/web/src/golang/src/slow_build
/usr/local/Cellar/go/1.4.2/libexec/pkg/tool/darwin_amd64/6g -o $WORK/slow_build.a -trimpath $WORK -p slow_build -complete -D _/Users/ianic/work/web/src/golang/src/slow_build -I $WORK -I /Users/ianic/work/web/src/golang/pkg/darwin_amd64 -pack ./main.go
cd .
/usr/local/Cellar/go/1.4.2/libexec/pkg/tool/darwin_amd64/6l -o $WORK/slow_build/_obj/exe/a.out -L $WORK -L /Users/ianic/work/web/src/golang/pkg/darwin_amd64 -extld=clang++ $WORK/slow_build.a
mv $WORK/slow_build/_obj/exe/a.out slow_build
real 0m1.075s
user 0m0.964s
sys 0m0.108s
**1.5 output for: time go build -v -x**
WORK=/var/folders/dx/p9b300dd27dcbqcrz0h9kn800000gn/T/go-build145414805
slow_build
mkdir -p $WORK/slow_build/_obj/
mkdir -p $WORK/slow_build/_obj/exe/
cd /Users/ianic/work/web/src/golang/src/slow_build
/usr/local/Cellar/go/1.5/libexec/pkg/tool/darwin_amd64/compile -o $WORK/slow_build.a -trimpath $WORK -p main -complete -buildid 14ca37ec220e553c7bf9829fe97751df5a041f5f -D _/Users/ianic/work/web/src/golang/src/slow_build -I $WORK -I /Users/ianic/work/web/src/golang/pkg/darwin_amd64 -pack ./main.go
cd .
/usr/local/Cellar/go/1.5/libexec/pkg/tool/darwin_amd64/link -o $WORK/slow_build/_obj/exe/a.out -L $WORK -L /Users/ianic/work/web/src/golang/pkg/darwin_amd64 -extld=clang++ -buildmode=exe -buildid=14ca37ec220e553c7bf9829fe97751df5a041f5f $WORK/slow_build.a
mv $WORK/slow_build/_obj/exe/a.out slow_build
real 0m16.816s
user 0m16.225s
sys 0m1.226s
| help wanted,ToolSpeed,NeedsFix,early-in-cycle | medium | Major |
102,520,642 | youtube-dl | [req] egghead.io | technology list: https://egghead.io/technologies/angular2
series list: https://egghead.io/series/react-flux-architecture
search results list: https://egghead.io/search?q=testing
video page: https://egghead.io/lessons/react-development-environment-setup
| site-support-request | medium | Critical |
102,556,643 | go | x/mobile/cmd/gomobile: INSTALL_FAILED_INVALID_APK when dashes in package | Hi,
It seems that it is not possible to build an Android application from Go source code if there is a dash (`-`) somewhere in the Go package name. Since my organization on GitHub has a dash inside, this means I cannot build any Android applications whatsoever.
You can easily test this with the currently available Go 1.5 release version.
1. Take the `golang.org/x/mobile/example/basic` example application and move it to the `golang.org/x/mobile/example/basic-demo` folder.
2. Run `gomobile build -target=android golang.org/x/mobile/example/basic-demo` command.
3. Run the `adb install basic-demo` command.
You will get the following error:
```
6927 KB/s (3293500 bytes in 0.464s)
pkg: /data/local/tmp/basic-demo.apk
Failure [INSTALL_FAILED_INVALID_APK]
```
Now, you might ask me why didn't I use the `gomobile install` command instead. There seems to be another issue as well. If you run the `gomobile install` command, you will not receive any output to the terminal at all. Furthermore, you will even get an exit code of `0`. It seems that the `gomobile` command hides any issues coming from the `adb` command.
My best guess is that somewhere in the manifest you place the package name. At least for Java based applications, I know that they don't allow dashes in the package names and you are advised to substitute those with underscores. ([http://stackoverflow.com/questions/3179216/what-is-the-convention-for-word-separator-in-java-package-names](http://stackoverflow.com/questions/3179216/what-is-the-convention-for-word-separator-in-java-package-names)) I guess a similar validation is performed even if the application is not Java based. This is just a hunch, I might be totally off base.
| mobile | low | Critical |
102,638,258 | go | cmd/go: cgo flags drops quotes. | There doesn't seem to be a way to pass a quoted value with cgo flags.
``` go
package libsass
/*
#cgo CFLAGS: -D VERSION="3.2.4-0ae11a4"
char* version(void) {
return VERSION;
}
*/
import "C"
func Version() string {
return C.GoString(C.version())
}
```
``` sh
$ go build -x -v
WORK=/tmp/go-build030587731
test/cgoflag
mkdir -p $WORK/test/cgoflag/_obj/
mkdir -p $WORK/test/cgoflag/_obj/exe/
cd /home/ome/go/src/test/cgoflag
CGO_LDFLAGS="-g" "-O2" /usr/local/go/pkg/tool/linux_amd64/cgo -objdir $WORK/test/cgoflag/_obj/ -importpath test/cgoflag -- -I $WORK/test/cgoflag/_obj/ -D VERSION=3.2.4-0ae11a4 main.go
# test/cgoflag
./main.go: In function 'version':
<command-line>:0:9: error: too many decimal points in number
./main.go:8:10: note: in expansion of macro 'VERSION'
return VERSION;
^
<command-line>:0:15: error: invalid suffix "ae11a4" on integer constant
./main.go:8:10: note: in expansion of macro 'VERSION'
return VERSION;
```
I have also tried `'"3.2.4-0ae11a4"'` and some other variations, they all yield:
`can't load package: package test/cgoflag: /home/ome/go/src/test/cgoflag/main.go: malformed #cgo argument: VERSION="3.2.4-0ae11a4"`
| NeedsInvestigation | low | Critical |
102,664,089 | youtube-dl | [Weibo] Unsupported URL | - **URL:** http://video.weibo.com/show?fid=1034:7ffc79010fec610ed9a7dc35bd288062
- **Error:** Unsupported URL: http://video.weibo.com/show?fid=1034:7ffc79010fec610ed9a7dc35bd288062
- **Platform:** openSUSE 13.1, x86_64, XFCE
- **Version:** 2015.08.16.1
- **Video player:** Custom
Oddly, there's supposed to be Weibo support, it seems (#1039). Not sure why it doesn't work here. Different video provider, maybe?
```
sven@linux-etoq:~/archive/shandong-explosion> youtube-dl -v http://video.weibo.com/show?fid=1034:7ffc79010fec610ed9a7dc35bd288062
[debug] System config: []
[debug] User config: []
[debug] Command-line args: [u'-v', u'http://video.weibo.com/show?fid=1034:7ffc79010fec610ed9a7dc35bd288062']
[debug] Encodings: locale UTF-8, fs UTF-8, out UTF-8, pref UTF-8
[debug] youtube-dl version 2015.08.16.1
[debug] Python version 2.7.6 - Linux-3.11.10-29-desktop-x86_64-with-SuSE-13.1-x86_64
[debug] exe versions: ffmpeg 2.7.2, ffprobe 2.7.2, rtmpdump 2.4
[debug] Proxy map: {}
[generic] show?fid=1034:7ffc79010fec610ed9a7dc35bd288062: Requesting header
WARNING: Falling back on generic information extractor.
[generic] show?fid=1034:7ffc79010fec610ed9a7dc35bd288062: Downloading webpage
[generic] show?fid=1034:7ffc79010fec610ed9a7dc35bd288062: Extracting information
ERROR: Unsupported URL: http://video.weibo.com/show?fid=1034:7ffc79010fec610ed9a7dc35bd288062
Traceback (most recent call last):
File "/usr/lib/python2.7/site-packages/youtube_dl/extractor/generic.py", line 1211, in _real_extract
doc = parse_xml(webpage)
File "/usr/lib/python2.7/site-packages/youtube_dl/utils.py", line 1594, in parse_xml
tree = xml.etree.ElementTree.XML(s.encode('utf-8'), **kwargs)
File "/usr/lib64/python2.7/xml/etree/ElementTree.py", line 1300, in XML
parser.feed(text)
File "/usr/lib64/python2.7/xml/etree/ElementTree.py", line 1642, in feed
self._raiseerror(v)
File "/usr/lib64/python2.7/xml/etree/ElementTree.py", line 1506, in _raiseerror
raise err
ParseError: syntax error: line 1, column 0
Traceback (most recent call last):
File "/usr/lib/python2.7/site-packages/youtube_dl/YoutubeDL.py", line 655, in extract_info
ie_result = ie.extract(url)
File "/usr/lib/python2.7/site-packages/youtube_dl/extractor/common.py", line 287, in extract
return self._real_extract(url)
File "/usr/lib/python2.7/site-packages/youtube_dl/extractor/generic.py", line 1804, in _real_extract
raise UnsupportedError(url)
UnsupportedError: Unsupported URL: http://video.weibo.com/show?fid=1034:7ffc79010fec610ed9a7dc35bd288062
```
| site-support-request | low | Critical |
102,713,176 | opencv | WinRT: imread/ imwrite not working properly in Visual Studio 2015 C++ Universal Windows Application | I wrote a code like this and `imread` only works when the path is in the ms-app folder.
I've changed the permission of the application in AppXManifest to make it able to access My Photo but doesn't help either.
The OpenCV branch I'm using is https://github.com/Microsoft/opencv/tree/vs2015-samples
I'm using Visual Studio ver. 14.0.23107.0 D14REL running on Windows 10.10240.
```
create_task(openPicker->PickSingleFileAsync()).then([this](StorageFile^ file)
{
if (file != nullptr)
{
**img = cv::imread(ConvertPath(file->Path));**
if (!img.empty())
{
textBlock->Text = "open Successful!";
cv::Mat tmp;
cv::cvtColor(img, tmp, cv::COLOR_BGR2BGRA);
UpdateImage(tmp);
}
else
{
textBlock->Text = "open " + file->Path + " failed!";
}
}
});
```
**Update**
If I load the files in WinRT ways, works.
```
create_task(openPicker->PickSingleFileAsync()).then([this](StorageFile^ file)
{
if (file != nullptr)
{
//WinRT way
RandomAccessStreamReference^ streamRef = RandomAccessStreamReference::CreateFromFile(file);
task<IRandomAccessStreamWithContentType^>(streamRef->OpenReadAsync()).
then([](task<IRandomAccessStreamWithContentType^> thisTask)
{
IRandomAccessStreamWithContentType^ fileStream = thisTask.get();
return BitmapDecoder::CreateAsync(fileStream);
}).
then([](task<BitmapDecoder^> thisTask)
{
BitmapDecoder^ decoder = thisTask.get();
return decoder->GetFrameAsync(0);
}).
then([this](task<BitmapFrame^> thisTask)
{
BitmapFrame^ frame = thisTask.get();
// Save some information as fields
frameWidth = frame->PixelWidth;
frameHeight = frame->PixelHeight;
return frame->GetPixelDataAsync();
}).
then([this](task<PixelDataProvider^> thisTask)
{
PixelDataProvider^ pixelProvider = thisTask.get();
Platform::Array<byte>^ srcPixels = pixelProvider->DetachPixelData();
img = cv::Mat(frameHeight, frameWidth, CV_8UC4);
memcpy(img.data, srcPixels->Data, 4 * frameWidth*frameHeight);
if (!img.empty())
{
textBlock->Text = "open Successful!";
UpdateImage(img);
}
else
{
textBlock->Text = "open failed!";
}
});
}
});
```
**Update 2**
The code I use in the sample of `imwrite` doesn't work either. No file is output.
https://github.com/Microsoft/opencv/blob/master/samples/winrt/OcvImageProcessing/OcvImageProcessing/MainPage.xaml.cpp
```
StorageFolder^ localFolderRT = ApplicationData::Current->LocalFolder;
cv::String localFile = ConvertPath(ApplicationData::Current->LocalFolder->Path) + "\\Lena.png";
return cv::imwrite(localFile, img);
```
Original thread: https://github.com/microsoft/opencv/issues/57
| bug,priority: low,affected: 3.4,platform: winrt/uwp | low | Critical |
102,778,347 | go | net/http: implementing size capped connection pools | Hi,
This issue came up a few times already on golang-nuts. We could use an upper bound on the HTTP connection pool (active + idle) size and either block the caller until a connection becomes available, or return an error.
One of the suggested approaches is to provide a custom Transport that does the bookkeeping inside Dial(). That works only partially, as after the HTTP Client call (Do() or any of the verb helpers) returns there is no way to get a reference to the underlying connection and return that to the pool.
net/httputil contains a ClientConn type that seems to have some potential, but it is considered old (and probably will disappear soon). Even so, as far as I can tell the only way to grab the underlying TCP connection is to hijack it, which in turn forces one to handle the entire HTTP logic themselves.
A naive approach would be to pool HTTP requests instead, but that is not really the desired approach.
Thanks for reading!
| NeedsInvestigation | low | Critical |
102,949,671 | youtube-dl | [bilibili] Add support for authentication | ```
youtube-dl http://www.bilibili.com/video/av2757963/ --verbose --dump-pages
[debug] System config: []
[debug] User config: [u'-f', u'bestvideo+bestaudio/best']
[debug] Command-line args: [u'http://www.bilibili.com/video/av2757963/', u'--verbose', u'--dump-pages']
[debug] Encodings: locale UTF-8, fs utf-8, out UTF-8, pref UTF-8
[debug] youtube-dl version 2015.08.23
[debug] Python version 2.7.10 - Darwin-14.5.0-x86_64-i386-64bit
[debug] exe versions: ffmpeg 2.7.2, ffprobe 2.7.2
[debug] Proxy map: {}
[BiliBili] 2757963: Downloading webpage
[BiliBili] Dumping request to http://www.bilibili.com/video/av2757963/
<!DOCTYPE html>
<html xmlns="http://www.w3.org/1999/xhtml">
	<head>
		
		<meta http-equiv="Content-Type" content="text/html; charset=utf-8">
		<title>哔哩哔哩弹幕视频网 - ( ゜- ゜)つロ  乾杯~  - bilibili</title>

		
		<meta name="description" content="bilibili是国内知名的视频弹幕网站，这里有最及时的动漫新番，最棒的ACG氛围，最有创意的Up主。大家可以在这里找到许多欢乐。">
		<meta name="keywords" content="B站,弹幕,字幕,AMV,MAD,MTV,ANIME,动漫,动漫音乐,游戏,游戏解说,ACG,galgame,动画,番组,新番,初音,洛天依,vocaloid">
		
		<meta name="renderer" content="webkit">
		<meta http-equiv="X-UA-Compatible" content="IE=edge">
		<link rel="shortcut icon" href="http://static.hdslb.com/images/favicon.ico">
		<link rel="search" type="application/opensearchdescription+xml" href="http://static.hdslb.com/opensearch.xml" title="哔哩哔哩" />
		
		<link rel="stylesheet" href="http://static.hdslb.com/images/jquery-ui/custom/jquery-ui.css" type="text/css">
		<link rel="stylesheet" href="http://static.hdslb.com/css/new_z2.css" type="text/css" />
		
		<script type="text/javascript" src="http://static.hdslb.com/js/jquery.min.js"></script>
		<script type="text/javascript" src="http://static.hdslb.com/js/jquery-ui.min.js"></script>
		<script type="text/javascript" src="http://static.hdslb.com/js/base.core.v3.js"></script>
		
		
		
	</head>

	<body>
		
		<script type="text/javascript">biliAdjust(); $(function(){ biliAdjust(true); $(window).resize(biliAdjust); });</script>
		
		<div class="float_window"><div class="close"></div><div id="float_window"></div></div>
		
		<div class="z_top_container">
			<div class="z_top">
				<div class="z_header">
					<div class="z_top_nav">
						<ul>
	<li class="home">
		<a class="i-link" href="http://www.bilibili.com/">主站</a>
	</li>
		<li class="hbili"><a class="i-link" href="http://h.bilibili.com/" title="画友">画友<em class="new"></em></a></li>
	<li>
		<a class="i-link" href="http://game.bilibili.com/" title="游戏中心">游戏中心<em class="beta"></em></a>
		<div class="i_div game">
	<div class="i_menu_bg_t"></div>
	
	<div class="hot">
 		<div class="game_img_box">
			<a target="_blank" href="http://acg.tv/uCk"><div class="img" style="background: url(http://i2.hdslb.com/u_user/2d8aed58ca483118433349749e106ec3.jpg);"></div><p><b class="t">MA2</b></p></a>
		</div>
		<div class="game_img_box">
			<a target="_blank" href="http://acg.tv/uCl"><div class="img" style="background: url(http://i2.hdslb.com/u_user/game/xx/img/70.jpg);"></div><p><b class="t">小熊料理</b></p></a>
		</div>
		<!-- <div class="game_img_box">
			<a target="_blank" href="http://tr.biligame.com/"><div class="img" style="background: url(http://static.hdslb.com/gamecms/img/nt/ic.jpg);"></div><p><b class="t">那年那兔</b></p></a>
		</div> -->
		<div class="game_img_box">
			<a target="_blank" href="http://acg.tv/uCm"><div class="img" style="background: url(http://i1.hdslb.com/u_user/game/li/img/li-icon.jpg);"></div><p><b class="t">鲤</b></p></a>
		</div>
		<!-- <div class="game_img_box">
			<a target="_blank" href="http://hhth.biligame.com/"><div class="img" style="background: url(http://i1.hdslb.com/u_user/game/hhth/img/icon72.png);"></div><p><b class="t">坏坏童话</b></p></a>
		</div> -->
		<!-- <div class="game_img_box">
			<a target="_blank" href=" http://yysjs.biligame.com "><div class="img" style="background: url(http://i2.hdslb.com/u_user/game/yysjs/img/icon0519/icon-yysjs.jpg);"></div><p><b class="t">悠悠世界树</b></p></a>
		</div> -->
		<!-- <div class="game_img_box">
			<a target="_blank" href="http://sj2.biligame.com/"><div class="img" style="background: url(http://i1.hdslb.com/u_user/game/2015world2/img/W2-ICON_80.png);"></div><p><b class="t">世界2</b></p></a>
		</div> -->
		
		<div class="game_img_box">
			<a target="_blank" href="http://acg.tv/uCn"><div class="img" style="background: url(http://i1.hdslb.com/u_user/game/hxzj_index/common-icons.png);"></div><p><b class="t">幻想战姬</b></p></a>
		</div>
		<!-- <div class="game_img_box">
			<a target="_blank" href="http://bmjh.biligame.com/"><div class="img" style="background: url(http://i1.hdslb.com/u_user/game/whitecat/img/wcat-icon.jpg);"></div><p><b class="t">白猫计划</b></p></a>
		</div> -->
		<div class="game_img_box">
			<a target="_blank" href="http://acg.tv/uCo"><div class="img" style="background: url(http://i2.hdslb.com/u_user/activities/mihoyo/img/hot3-2.png);"></div><p><b class="t">崩坏学园2</b></p></a>
		</div>
		<!-- <div class="game_img_box">
			<a target="_blank" href="http://gz.biligame.com/"><div class="img" style="background: url(http://i2.hdslb.com/u_user/game/gz/img/gz0603.jpg);"></div><p><b class="t">我家公主最可爱</b></p></a>
		</div> -->
		<div class="game_img_box">
			<a target="_blank" href="http://acg.tv/uCp"><div class="img" style="background: url(http://i2.hdslb.com/u_user/game/xsqst/img/xsqst-icon.jpg);"></div><p><b class="t">像素骑士团</b></p></a>
		</div>
	</div>
	<div class="all">
		<b>新游预告:</b>
		    <!-- <a href="http://cb.biligame.com/" target="_blank">小熊料理</a> -->
		    <a href="http://acg.tv/uCq" target="_blank">100王子</a>
		    <!-- <a href="http://www.bilibili.com/video/av2641452/" target="_blank">鲤</a>
		    <a href="http://ma.biligame.com/" target="_blank">乖离性百万亚瑟王</a> -->
		<!-- <a href="http://teos2.biligame.com/" target="_blank">崩坏学园2</a> -->
		<!-- <a href="http://xwy.biligame.com/" target="_blank">侠物语</a> -->
		
		<!-- <a href="http://www.bilibili.com/html/ma_pc.html" target="_blank">百万亚瑟王</a> -->
		<!-- <a href="http://szr.biligame.com/" target="_blank">神之刃</a> -->
		<!--<a href="http://www.bilibili.com/html/game/guardiancross_pc.html" target="_blank">魔物狩猎者</a>-->
		<!-- <a href="http://game.bilibili.com/flappy-bilibili/index_v2.html" target="_blank">飞翔小电视</a> -->
		<!-- <a href="http://game.bilibili.com/2048.html" target="_blank">2048</a> -->
		
		<!-- <a href="http://yzr.biligame.com/" target="_blank">影之刃</a> -->
		<!-- <a href="http://qwmx.biligame.com/" target="_blank">蔷薇梦想</a> -->
		<!-- <a href="http://mlk.biligame.com" target="_blank">梅露可物语</a> -->
	</div>
</div>	</li>
		
	<li class="live"><a class="i-link" target="_blank" href="http://live.bilibili.com" title="直播">直播<em class="beta"></em></a></li>
	<li class="no_border"><a class="i-link" target="_blank" href="http://app.bilibili.com" title="手机端">手机端</a></li>
	<li class="yoo"><a class="i-link" target="_blank" href="http://yoo.bilibili.com/index.html" title="日本游">日本游<em class="new"></em></a></li>
</ul>					</div>
					<div class="uns_box">
	<ul class="menu">
		<li id="i_menu_profile_btn" guest="no" class="u-i i_user" i_menu="#i_menu_profile">
			<a class="i-link" href="http://member.bilibili.com/" target="_blank">
				<img class="i_face">
			</a>
			<div id="i_menu_profile" class="i_menu">
				<div class="i_menu_bg_t"></div>
				<div class="info clearfix"><div class="uname"></div><div class="coin"></div></div>
				<ul class="member-menu">
					<li><a href="http://space.bilibili.com/" target="_blank" class="member">个人空间</a></li>
					<li><a target="_blank" href="https://account.bilibili.com/" class="account">账户中心</a></li>
				</ul>
				<div class="member-bottom">
					<a class="logout" href="https://account.bilibili.com/login?act=exit">退出</a>
				</div>
			</div>
		</li>
		<li id="i_menu_community_msg_btn" guest="no" i_menu="community_msg" class="u-i">
            <a class="i-link" href="http://message.bilibili.com" target="_blank">消息</a>
        </li>
		<li id="i_menu_msg_btn" guest="no" i_menu="#dyn_wnd" class="u-i">
			<div class="num" id="dynamic_num_total"></div>
			<a class="i-link" href="http://www.bilibili.com/account/dynamic" target="_blank">动态</a>
		</li>
		<li id="i_menu_fav_btn" guest="no" i_menu="#i_menu_fav" class="u-i">
			<a class="i-link" href="http://space.bilibili.com/#!/fav" target="_blank">收藏夹</a>
		</li>
		<li id="i_menu_history_btn" guest="no" class="u-i">
			<a class="i-link" href="http://www.bilibili.com/account/history">历史</a>
		</li>
		<li guest="no" class="u-i b-post">
			<a class="i-link" href="http://member.bilibili.com/video_submit.html?tpl=upload" target="_blank">投稿</a><ul class="s-menu"><li><a href="http://member.bilibili.com/#video_manage" target="_blank"><i class="icon-mng"></i><em>视频管理</em></a></li></ul>
		</li>
		<li id="i_menu_login_btn" guest="yes" class="u-i">
			<ul class="index-login">
				<li class="b-btn"><a class="icon-qq" href="https://account.bilibili.com/login?sns=qq"></a></li>
				<li class="b-btn"><a class="icon-sina" href="https://account.bilibili.com/login?sns=weibo"></a></li>
				<li class="b-btn login"><a class="login" href="https://account.bilibili.com/login">登录</a></li>
			</ul>
		</li>
		<li id="i_menu_register_btn" guest="yes" class="u-i b-reg">
			<a class="i-link" href="/register">注册</a>
		</li>
	</ul>
</div>				</div>
			</div>
		</div>
		<div class="header">
			<div class="h-center">
				<a href="/" class="logo"></a>
				<a href="http://www.bilibili.com/random" target="_blank" title="随便找个视频看看?" class="lsb"></a>
			</div>
			<div class="num">
				<div class="menu-wrapper">			
					<ul>
	<li class="m-i home"><a class="i-link" href="/">首页</a></li>
	<!-- 动画 -->
	<li class="m-i ">
		<a class="i-link" href="/video/douga.html"><em>动画</em></a>
		<ul class="i_num">
			<li><a href="/video/douga-mad-1.html"><b>MAD·AMV<em></em></b></a></li>
			<li><a href="/video/douga-mmd-1.html"><b>MMD·3D<em></em></b></a></li>
			<li><a href="/video/douga-voice-1.html"><b>动画短片<em></em></b></a></li>
			<li><a href="/video/douga-else-1.html"><b>综合<em></em></b></a></li>
		</ul>
		<span class="addnew_1">--</span>
	</li>
	<!-- 番剧 -->
	<li class="m-i ">
		<a class="i-link" href="/video/bangumi.html"><em>番剧</em></a>
		<ul class="i_num">
			<li><a href="/video/bangumi-two-1.html"><b>连载动画<em></em></b></a></li>
			<li><a href="/video/part-twoelement-1.html"><b>完结动画<em></em></b></a></li>
			<li><a href="/video/douga-else-information-1.html"><b>资讯<em></em></b></a></li>
			<li><a href="/video/bagumi_offical_1.html"><b>官方延伸<em></em></b></a></li>
			<li><a href="/video/bangumi_chinese_1.html"><b>国产动画<em></em></b></a></li>
			<li><a href="/list/b--a--t----d---1.html"><b>新番Index<em></em></b></a></li>
		</ul>
		<span class="addnew_13">--</span>
	</li>
	<!-- 音乐 -->
	<li class="m-i ">
		<a class="i-link" href="/video/music.html"><em>音乐</em></a>
		<ul class="i_num">
			<li><a href="/video/music-Cover-1.html"><b>翻唱<em></em></b></a></li>
			<li><a href="/video/music-vocaloid-1.html"><b>VOCALOID·UTAU<em></em></b></a></li>
			<li><a href="/video/music-perform-1.html"><b>演奏<em></em></b></a></li>
			<li><a href="/video/music-coordinate-1.html"><b>三次元音乐<em></em></b></a></li>
			<li><a href="/video/music-video-1.html"><b>同人音乐<em></em></b></a></li>
			<li><a href="/video/music-oped-1.html"><b>OP/ED/OST<em></em></b></a></li>
			<li><a href="/video/music-collection-1.html"><b>音乐选集<em></em></b></a></li>
		</ul>
		<span class="addnew_3">--</span>
	</li>
	<!-- 舞蹈 -->
	<li class="m-i ">
		<a class="i-link" href="/video/dance.html"><em>舞蹈</em></a>
		<ul class="i_num">
			<li><a href="/video/dance-1.html"><b>ACG相关舞蹈<em></em></b></a></li>
			<li><a href="/video/three-dimension-dance-1.html"><b>三次元舞蹈<em></em></b></a></li>
		</ul>
		<span class="addnew_129">--</span>
	</li>
	<!-- 游戏 -->
	<li class="m-i ">
		<a class="i-link" href="/video/game.html"><em>游戏</em></a>
		<ul class="i_num">
			<li><a href="/video/game-video-1.html"><b>单机联机<em></em></b></a></li>
			<li><a href="/video/game-ctary-network-1.html"><b>网络游戏<em></em></b></a></li>
			<li><a href="/video/game-fight-1.html"><b>电子竞技<em></em></b></a></li>
			<li><a href="/video/music-game-1.html"><b>音游<em></em></b></a></li>
			<li><a href="/video/game-mugen-1.html"><b>Mugen<em></em></b></a></li>
			<li><a href="/video/gmv-1.html"><b>GMV<em></em></b></a></li>
		</ul>
		<span class="addnew_4">--</span>
	</li>
	<!-- 科技 -->
	<li class="m-i ">
		<a class="i-link" href="/video/technology.html"><em>科技</em></a>
		<ul class="i_num">
			<li><a href="/video/tech-popular-science-1.html"><b>纪录片<em></em></b></a></li>
			<li><a href="/video/tech-fun-1.html"><b>趣味科普人文<em></em></b></a></li>
			<li><a href="/video/tech-wild-1.html"><b>野生技术协会<em></em></b></a></li>
			<li><a href="/video/speech-course-1.html"><b>演讲·公开课<em></em></b></a></li>
			<li><a href="/video/tech-future-military-1.html"><b>军事<em></em></b></a></li>
			<li><a href="/video/tech-future-digital-1.html"><b>数码<em></em></b></a></li>
			<li><a href="/video/tech-future-other-1.html"><b>机械<em></em></b></a></li>
		</ul>
		<span class="addnew_36">--</span>
	</li>
	<!-- 娱乐 -->
	<li class="m-i ">
		<a class="i-link" href="/video/ent.html"><em>娱乐</em></a>
		<ul class="i_num">
			<li><a href="/video/ent_funny_1.html"><b>搞笑<em></em></b></a></li>
			<li><a href="/video/ent-life-1.html"><b>生活<em></em></b></a></li>
			<li><a href="/video/ent-animal-1.html"><b>动物圈<em></em></b></a></li>
			<li><a href="/video/ent-food-1.html"><b>美食圈<em></em></b></a></li>
			<li><a href="/video/ent-variety-1.html"><b>综艺<em></em></b></a></li>
			<li><a href="/video/ent-circle-1.html"><b>娱乐圈<em></em></b></a></li>
			<li><a href="/video/ent-korea-1.html"><b>Korea相关<em></em></b></a></li>
		</ul>
		<span class="addnew_5">--</span>
	</li>
	<!-- 鬼畜 -->
	<li class="m-i ">
		<a class="i-link" href="/video/kichiku.html"><em>鬼畜</em></a>
		<ul class="i_num">
			<li><a href="/video/douga-kichiku-1.html"><b>二次元鬼畜<em></em></b></a></li>
			<li><a href="/video/ent-Kichiku-1.html"><b>三次元鬼畜<em></em></b></a></li>
			<li><a href="/video/kichiku-manual_vocaloid-1.html"><b>人力VOCALOID<em></em></b></a></li>
			<li><a href="/video/kichiku-course-1.html"><b>教程演示<em></em></b></a></li>
		</ul>
		<span class="addnew_119">--</span>
	</li>
	<!-- 电影 -->
	<li class="m-i ">
		<a class="i-link" href="/video/movie.html"><em>电影</em></a>
		<ul class="i_num">
			<li><a href="/video/movie_west_1.html"><b>欧美电影<em></em></b></a></li>
			<li><a href="/video/movie_japan_1.html"><b>日本电影<em></em></b></a></li>
			<li><a href="/video/movie_chinese_1.html"><b>国产电影<em></em></b></a></li>
			<li><a href="/video/movie-movie-1.html"><b>其他国家<em></em></b></a></li>
			<li><a href="/video/tv-micromovie-1.html"><b>短片<em></em></b></a></li>
			<li><a href="/video/movie-presentation-1.html"><b>电影相关<em></em></b></a></li>
		</ul>
		<span class="addnew_23">--</span>
	</li>
	<!-- 电视剧 -->
	<li class="m-i ">
		<a class="i-link" href="/video/teleplay.html"><em>电视剧</em></a>
		<ul class="i_num">
			<li><a href="/video/soap-three-1.html"><b>连载剧集<em></em></b></a></li>
			<li><a href="/video/tv-drama-1.html"><b>完结剧集<em></em></b></a></li>
			<li><a href="/video/tv-sfx-1.html"><b>特摄·布袋<em></em></b></a></li>
			<li><a href="/video/tv-presentation-1.html"><b>电视剧相关<em></em></b></a></li>
		</ul>
		<span class="addnew_11">--</span>
	</li>
	
	<li class="m-i rank"><a class="i-link" href="http://www.bilibili.com/ranking">排行</a></li>
</ul>					<div class="search">
						<form action="http://www.bilibili.com/search" name="formsearch" id="searchform" target="_blank" _lpchecked="1">
						<input name="keyword" type="text" style="border:none;box-shadow:none;" class="search-keyword ui-autocomplete-input" id="search-keyword" autocomplete="off" role="textbox" aria-autocomplete="list" aria-haspopup="true" accesskey="s" x-webkit-speech="" x-webkit-grammar="builtin:translate">
						<input name="orderby" type="hidden" value="" />
						<button type="submit" name="formsubmit" class="search-submit"></button>
						</form>
					</div>
				</div>
			</div>
		</div>
            
		<div class="z">

			
<style type="text/css">
    .widescreen .z,.z{ width:100%; }
    .footer{ margin-top:0; }
    .z-msg{ background-color:#edf1f8; }
    .z-msg .z-msg-box{ position:relative;margin:0 auto;text-align:center;padding:23px 0px 125px 0;width:980px;height: 364px; }
    .z-msg #turnToLogin{ background-color:#f489ad;position:absolute;left:588px;top:280px;border-radius:5px;line-height:28px;padding:0 22px 0 27px;color:#edf1f8; }
    .z-msg #turnToLogin:hover{ background-color:#ffa3c2; }
    .z-msg .msgbox{
        position: absolute;
        right: 185px;
        top: 145px;
        width: 300px;
        text-align: center;
        line-height: 16px;
    }
    .z-msg .msgbox .title{
        line-height: 30px;
        text-align: center;
        font-size: 24px;
        color: #4d73b1;
        margin-bottom: 15px;
    }
    .z-msg .msgbox .content{
        display: inline-block; *display: inline; *zoom:1;
        text-align: left;
    }
    .z-msg .bg{
        margin: 0 auto;
        width: 773px;
        height: 364px;
        background: url(http://static.hdslb.com/error/errorPage.png) no-repeat;
    }
</style>
<div class="z-msg">
    <div class="z-msg-box">
        <div class="bg"></div>
        <div class="msgbox"><div class="title"><b>好像不能看了...</b></div><div class="content">你没有权限浏览！可能　1.您<font color="red">无权访问</font>本页面　2.本页面已被<font color="red">删除</font><br /><br />您可以尝试 <a href="https://account.bilibili.com/login">登录</a> 或 <a href="https://account.bilibili.com/register/phone">注册帐号</a></div></div>
        
    </div>
</div>

			
		</div>
		
		<div class="footer">
	<div class="footer-cnt clearfix">
		<ul class="boston-postcards">
			<li>
				<a target="_blank" href="http://www.bilibili.com/html/aboutUs.html">关于我们</a><br />
				<a target="_blank" href="http://www.bilibili.com/html/contact.html">联系我们</a><br />
			    <a target="_blank"  href="http://www.bilibili.com/html/join.html">加入我们</a><br />
			    <a target="_blank"  href="http://www.bilibili.com/html/friends-links.html">友情链接</a><br />
				<a target="_blank"  target="_blank" href="http://weibo.com/bilibiliweb">新浪微博</a> | 
				<a id="weixin">官方微信
					<div class="wx-ewm-box" id="wx">
					  <div class="wx-ewm">
					    <div class="wx-arrow">
					    </div>
					  </div>
					</div>
			    </a>
			</li>
			<li>
				<a target="_blank" href="http://h.bilibili.com/">画友</a><br />
				<a target="_blank" href="http://game.bilibili.com/">游戏中心</a><br/>
				<a target="_blank"  href="http://www.bilibili.com/lottery">活动专题页</a><br />
				<a target="_blank"  href="http://h.bilibili.com/wallpaper?action=list">壁纸站</a><br />
				<a target="_blank"  href="https://account.bilibili.com/answer/addq">分院帽计划</a>
			</li>
			<li>
				<a target="_blank"  href="http://www.bilibili.com/html/help.html">帮助中心</a><br />
				<a target="_blank"  href="http://www.bilibili.com/html/userdel_list.html">补档认领处</a><br />
				<a target="_blank"  href="http://www.bilibili.com/video/av120040/">高级弹幕</a><br />
				<a target="_blank"  target="_blank" href="http://www.bilibili.com/html/copyright.html">侵权申诉</a><br/>
				<a target="_blank"  href="http://www.bilibili.com/html/duty.html">资源免责声明</a>
			</li>
			<li>
			    <a target="_blank" href="http://bilibili2233.taobao.com/">bilibili官方淘宝店铺</a><br />
			    <a target="_blank"  href="http://www.bilibili.com/html/cele.html">名人堂</a><br />
	            <a target="_blank" href="http://comic.qq.com/news/">资讯</a><br />
	            <a target="_blank" href="http://bbs.biligame.com/forum.php?mod=forumdisplay&fid=44">建议和BUG汇报</a>
			</li>
		</ul><div class="partner">
			<div class="block left">
				<div class="partner-banner"></div>
			</div>
			<div class="block right" style="margin: 0 0 0 10px;">
	        <p>广播电视节目制作经营许可证：<span>（沪）字第1248号 </span></p>
	        <p>网络文化经营许可证：<span>沪网文[2013]0480-056号</span></p>
	    	<p>互联网ICP备案：<span>沪ICP备13002172号-3</span></p>
	    	<p>沪ICP证：<span>沪B2-20100043</span></p>
	    	<p>信息网络传播视听节目许可证：<span>0910417</span></p>
	        </div>
		</div>
	</div>
</div>		
		<script type="text/javascript">loadLoginStatus();</script>
		<script type="text/javascript" src="http://www.bilibili.com/online.js"></script>
<div style="display:none;">
<script type="text/javascript">
$(function(){
	$.ajax({ url: 'http://data.bilibili.com/rec.js',cache: true,dataType:'script'});
	$('<scr'+'ipt type="text/javascript" src="http://static.hdslb.com/js/bfd.js" charset="UTF-8"></scri'+'pt>').appendTo('body');	
	$('<scr'+'ipt type="text/javascript" src="http://s11.cnzz.com/stat.php?id=2724999&web_id=2724999" charset="UTF-8"></scri'+'pt>').appendTo('body');	
	$('<scr'+'ipt type="text/javascript" src="http://tajs.qq.com/stats?sId=9156259" charset="UTF-8"></scri'+'pt>').appendTo('body');	
});
</script>


</div>
	</body>
</html>

ERROR: Unable to extract video code; 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.
Traceback (most recent call last):
File "/usr/local/bin/youtube-dl/youtube_dl/YoutubeDL.py", line 655, in extract_info
ie_result = ie.extract(url)
File "/usr/local/bin/youtube-dl/youtube_dl/extractor/common.py", line 287, in extract
return self._real_extract(url)
File "/usr/local/bin/youtube-dl/youtube_dl/extractor/bilibili.py", line 54, in _real_extract
r'(?s)<div itemprop="video".*?>(.*?)</div>', webpage, 'video code')
File "/usr/local/bin/youtube-dl/youtube_dl/extractor/common.py", line 569, in _search_regex
raise RegexNotFoundError('Unable to extract %s' % _name)
RegexNotFoundError: Unable to extract video code; 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.
```
| request | low | Critical |
103,000,355 | neovim | Wrong indenting of compound literals. | Indenting of compound literals is broken in both neovim and vim when assignment and declaration are split.
I'm using:
```
NVIM 0.0.0-alpha+201508240435 (compiled Aug 24 2015 19:21:22)
Commit: 5066128d484ee8d0cb10ddd2260c3f5b47104b92
VIM - Vi IMproved 7.4 (2013 Aug 10, compiled Jul 11 2015 08:44:54)
```
But happens with every version of neovim / vim I've tried. All of this with empty rc files.
Here is how it looks:
```
struct Short {
int a, b, c;
};
// Note: trailing commas on the last element make no difference whatsoever.
void IndentingStructs() {
// Right
struct Short foo = {
1,
2,
3,
};
// Right
struct Short bar = {
.a = 1,
.b = 2,
.c = 3,
};
// Wrong
foo = (struct Short){
1,
2,
3,
};
// Wrong
bar = (struct Short){
.a = 1,
.b = 2,
.b = 3,
};
}
```
| bug-vim,syntax | low | Critical |
103,013,185 | opencv | cv::Point::homogeneous() | Analog to Eigen's [MatrixBase::homogeneous](http://eigen.tuxfamily.org/dox/classEigen_1_1MatrixBase.html#af3513a0abfc3875e6d1912b8bb7b28ac) I vote for a cv::Point::homogeneous and a cv::Point3_::homogeneous , i.e. a convenience method which returns the same point with one additional dimension set to 1 so you can quickly multiply the point with a homography or a projection matrix.
Before implementing this, there are several questions to discuss:
1. How to call the method? homogeneous() seems quite long, compared to the other functions. **hom() ?**
2. What should the method return? Point::homogeneous() could return Point3 but what does Point3::homogeneous() return? I would vote for a **cv::Matx** to keep the interface clean
3. Should this method be only in the point classes or also in Matx where it throws a compile error for all non-1-column matrices?
4. Should we also introduce an equivalent for [hnormalize](http://eigen.tuxfamily.org/dox/classEigen_1_1VectorwiseOp.html#a1162b67a2744d546a6d22ecf56f47131)?
| priority: normal,feature,category: core | low | Critical |
103,232,492 | youtube-dl | Site Support Request: bugs.co.kr | Is it possible to add support for this url and website:
http://music.bugs.co.kr/mvPlayer?mvId=196710
I got member of this website, but couldn't buy it and download it due to this video is free to watch for everyone.
It got many Korean show undisclosed cut that are not for sale and free to watch.
Any chance to add support for this website?
Sorry for any inconvenience. Thank you.
P.S. Non-member could only watch the video by length of 1 minute and member could watch it by full-length.
| site-support-request | low | Critical |
103,276,638 | opencv | In faceDetect.cpp Video input still be executed after it finish (infinite loop) | Hi ,
when trying to execute the code faceDetect.cpp with a video input instead of the camera itself , the display won't stop and when the video is finished : I get this warning (result:10030): GLib-GObject-WARNING **: Attempt to add property GtkSettings::gtk-label-select-on-focus after class was initialised and it will block sometimes or start again in the same windows continuously , the only way to stop it is to hit Q or ctrl+c but I want that the video stop automatically after finishing processing all frames in video
i execute the code by
./faceDetect --cascade="haarcascade_frontalface_alt.xml" --nested-cascade="haarcascade_eye_tree_eyeglasses.xml" --scale=1 vidTest1
Any hints :neckbeard: ?
poke : @avdmitry and @vpisarev :v:
| bug,priority: normal,category: samples | low | Minor |
103,296,085 | opencv | Unable to compile calib3d with cuda | I get the following compilation error when compiling the latest master WITH_CUDA=True.
```
[ 83%] Building CXX object modules/cudalegacy/CMakeFiles/opencv_cudalegacy.dir/src/calib3d.cpp.o
/home/sam/src/opencv/modules/cudalegacy/src/calib3d.cpp: In function ‘void cv::cuda::solvePnPRansac(const cv::Mat&, const cv::Mat&, const cv::Mat&, const cv::Mat&, cv::Mat&, cv::Mat&, bool, int, float, int, std::vector<int>*)’:
/home/sam/src/opencv/modules/cudalegacy/src/calib3d.cpp:232:90: error: ‘undistortPoints’ was not declared in this scope
undistortPoints(image, image_normalized, camera_mat, dist_coef, Mat(), eye_camera_mat);
^
/home/sam/src/opencv/modules/cudalegacy/src/calib3d.cpp:232:90: note: suggested alternative:
In file included from /home/sam/src/opencv/build/modules/cudalegacy/precomp.hpp:60:0:
/home/sam/src/opencv/modules/calib3d/include/opencv2/calib3d.hpp:1692:23: note: ‘cv::fisheye::undistortPoints’
CV_EXPORTS_W void undistortPoints(InputArray distorted, OutputArray undistorted,
^
/home/sam/src/opencv/modules/cudalegacy/src/calib3d.cpp:255:5: error: ‘minMaxLoc’ is not a member of ‘cv::cuda’
cuda::minMaxLoc(d_hypothesis_scores, NULL, &best_score, NULL, &best_idx);
^
/home/sam/src/opencv/modules/cudalegacy/src/calib3d.cpp:255:5: note: suggested alternatives:
In file included from /home/sam/src/opencv/modules/core/include/opencv2/core/cuda.hpp:51:0,
from /home/sam/src/opencv/modules/cudalegacy/include/opencv2/cudalegacy.hpp:46,
from /home/sam/src/opencv/build/modules/cudalegacy/precomp.hpp:50:
/home/sam/src/opencv/modules/core/include/opencv2/core.hpp:779:17: note: ‘cv::minMaxLoc’
CV_EXPORTS void minMaxLoc(const SparseMat& a, double* minVal,
^
/home/sam/src/opencv/modules/core/include/opencv2/core.hpp:779:17: note: ‘cv::minMaxLoc’
make[2]: *** [modules/cudalegacy/CMakeFiles/opencv_cudalegacy.dir/src/calib3d.cpp.o] Error 1
make[1]: *** [modules/cudalegacy/CMakeFiles/opencv_cudalegacy.dir/all] Error 2
make: *** [all] Error 2
```
The system is Ubuntu 14.04 with CUDA 7.0. The output of CMake is bellow
| bug,priority: normal,category: build/install,affected: 3.4,category: gpu/cuda (contrib) | low | Critical |
103,322,744 | TypeScript | Expose `--init` functionality in a tsserver command | We now have `--init` support to generating a default tsconfig.json file (see #4037). Editors would need to expose a functionality that uses this, possibly passing a set of additional defaults based on their scenario. we need to expose this as a Language Service API as well as a tsserver command. see @jrieken's comment in https://github.com/Microsoft/TypeScript/pull/4037#issuecomment-135056980
| Suggestion,Help Wanted | low | Minor |
103,357,204 | rust | [MSVC] `cl.exe` doesn't support static relocations with large address awareness. | See `relocation-model` from `run-make`.
| O-windows,C-enhancement,T-compiler,O-windows-msvc | low | Minor |
103,436,992 | java-design-patterns | Service configurator pattern | ### Description:
The Service Configurator design pattern allows an application to dynamically configure and manage a set of services without requiring a restart. This pattern is useful for applications that need to handle a variety of services with different configurations and lifecycles. Key elements of the Service Configurator pattern include service registration, dynamic configuration, lifecycle management, and service lookup.
### Main elements of the Service Configurator pattern:
1. **Service Registration:** Services are registered with a central registry or configuration manager.
2. **Dynamic Configuration:** Services can be configured dynamically at runtime, allowing for changes without restarting the application.
3. **Lifecycle Management:** The pattern manages the lifecycle of services, including initialization, configuration, reconfiguration, and shutdown.
4. **Service Lookup:** Services can be located and accessed by clients through a central lookup mechanism.
### References:
- [Pattern-Oriented Software Architecture Volume 4: A Pattern Language for Distributed Computing](https://en.wikipedia.org/wiki/Pattern-Oriented_Software_Architecture)
- [Dynamically Configuring Communication Services with the Service Configurator Pattern](http://www.dre.vanderbilt.edu/~schmidt/PDF/O-Service-Configurator.pdf)
### Acceptance Criteria:
1. Implement a Service Configurator class that handles the registration, configuration, and lifecycle management of services.
2. Provide example services that demonstrate dynamic configuration and lifecycle management.
3. Include unit tests to verify the correct implementation of the Service Configurator pattern and its components.
| info: help wanted,epic: pattern,type: feature | low | Minor |
103,437,284 | java-design-patterns | Acceptor-Connector pattern | **Description:**
The Acceptor-Connector design pattern decouples connection establishment from connection handling in networked systems. The pattern's main components are the Acceptor, Connector, and Handler. The Acceptor is responsible for listening for incoming connection requests and creating a new connection. The Connector actively initiates connections to a remote service. Once a connection is established, the Handler takes over to manage the communication between peers.
Implementing this pattern will improve the system’s scalability and flexibility by separating connection handling from business logic, allowing each component to be managed and extended independently.
**References:**
1. [Extension and Java Implementation of the Reactor−Acceptor−Connector Pattern Combination](https://hillside.net/plop/plop99/proceedings/Fernandez3/RACPattern.PDF)
2. [Sample implementation](https://code.google.com/a/eclipselabs.org/p/comp-4104-carleton/)
3. [Pattern-Oriented Software Architecture Volume 2: Patterns for Concurrent and Networked Objects](https://amzn.to/3V5uoKO)
**Acceptance Criteria:**
1. Implement the Acceptor component to handle incoming connection requests.
2. Implement the Connector component to initiate outgoing connections.
3. Implement the Handler component to manage established connections and business logic.
4. Provide unit tests for each component to ensure proper functionality.
5. Update documentation with usage examples and detailed descriptions of each component.
| info: help wanted,epic: pattern,type: feature | low | Minor |
103,561,097 | rust | Type inference problem with lifetimes and assoiated types | I am trying to get some lifetime polymorphism but I unable to do so because of some bugs (#24424, #23958). This time it seems that the type inference is not working.
``` rust
use std::ops::Range;
pub trait Iter {
type Type: Iterator;
}
pub trait T {
fn f<'a>(&'a self) -> <&'a Self as Iter>::Type where &'a Self: Iter;
}
pub struct S;
impl<'a> Iter for &'a S {
type Type = Range<usize>;
}
impl T for S {
fn f<'a>(&'a self) -> <&'a Self as Iter>::Type where &'a Self: Iter {
0usize..10
}
}
fn main() {}
```
gives
```
<anon>:19:9: 19:19 error: mismatched types:
expected `<&'a S as Iter>::Type`,
found `core::ops::Range<usize>`
(expected associated type,
found struct `core::ops::Range`) [E0308]
<anon>:19 0usize..10
^~~~~~~~~~
<anon>:19:9: 19:19 help: see the detailed explanation for E0308
error: aborting due to previous error
playpen: application terminated with error code 101
```
| A-lifetimes,P-low,A-associated-items,I-needs-decision,T-lang,C-bug | low | Critical |
103,632,573 | rust | Add an overview toggle to rustdoc lib output | Rustdoc libs should have an overview button of some sort.
As discussed [here](https://internals.rust-lang.org/t/rustdoc-mockup-for-including-more-structured-content/2540), one problem with the std docs is that you can't tell which part of an unfamiliar page is actually useful. For example, below is a mockup of what the main [regex page](http://doc.rust-lang.org/regex/regex/index.html) would look like with and without the types which can only be created indirectly. Clearly the top part is much more approachable to someone unfamiliar with this library.
This seems a somewhat reasonable thing to do although I'm not certain `expand/overview` are the best words or icons for this but it seems okay I guess. One alternative would be to only show indirectly created types where they're being returned (but that is more drastic and it seems it could be done after this was tried if this was really well liked).
cc @BurntSushi since this interested you
| T-rustdoc,C-feature-request | low | Minor |
103,801,476 | TypeScript | Improve error message for decorator type mismatches | In TS 1.5 this:
``` ts
//Relation.ts
export function Relation(cls: any) {
return function () {
console.log(cls);
}
}
//Foo.ts
import {Relation} from "./Relation";
export class Foo {
@Relation(Foo)
info: string;
}
```
Compiles without error. In TS 1.6, it reports the error:
```
error TS1240: Unable to resolve signature of property decorator when called as an expression.
Supplied parameters do not match any signature of call target.
```
On the decorator call in `Foo.ts`. This is a very cryptic message. What it wants you to do is change `Relation.ts` to read like this:
``` ts
export function Relation(cls: any): PropertyDecorator {
return function () {
console.log(cls);
}
}
```
or this:
``` ts
export function Relation(cls: any) {
return function (target: Object, propertyKey: string | symbol) {
console.log(cls);
}
}
```
It would be very useful for the error message to indicate what exactly is wrong and how it could be fixed. Something akin to
```
error TS1240: Expression type is not assignable to decorator type [[PropertyDecorator]]. Ensure [[@Relation(Foo)]] has a type assignable to [[PropertyDecorator]].
```
would be more useful in actually understanding what's wrong and why you need to change your code.
| Bug,Help Wanted,Good First Issue,Domain: Error Messages | medium | Critical |
103,816,001 | TypeScript | Improve support for Symbol.toPrimitive | Presently, TS limits index types to 'string', 'number', 'symbol', or 'any'. That's fine, but consider the following:
``` ts
class Foo {
[Symbol.toPrimitive](): number {
return 1;
}
}
var x = ['a','b','c'];
var f = new Foo();
x[f];
```
We give the error `an index expression argument must be of type ...` on the array access.
`Foo` is for all intents and purposes a wrapper around the number `1` (close to `new Number(1)`), yet it cannot be used to index an array, since `Foo` itself is not of type `number`, `string`, `symbol`, or `any`. We could do extra type-checking to see if an object has a user-defined `Symbol.toPrimitive` and consider its return types when checking type-coercing operations (after all, if you've opted in to using `Symbol.toPrimitve` it's probably part of your goal to gain this behavior), such as indexing or mathematics operators.
Building on this, we currently don't support the coercion of `Number` or `String` classes to their respective primitive types (which are a specialization of the above issue but which applies to ES5 and ES3 output). Presently one can't do this:
``` ts
function add(a: string, b: string): number {
var x = new Number(a);
var y = new Number(b);
return x + y;
}
add('10e4', '0x8');
```
Because since neither x nor y is the 'number' (or 'string') primitive type, the addition operation is not allowed. Though it's still valid js (and without the usage of the Number class, the addition would result in '10e40x8'!).
| Suggestion,Needs Proposal | medium | Critical |
103,817,894 | TypeScript | Poor error recovery in jsx elements | ``` jsx
var x = <div span=3 ></div>;
```
```
Error TS1005 '{' expected. TypeScriptHTMLApp6 c:\Users\mhegazy\Documents\Visual Studio 2015\Projects\TypeScriptHTMLApp6\TypeScriptHTMLApp6\file1.tsx 10
Error TS17002 Expected corresponding JSX closing tag for 'div'. TypeScriptHTMLApp6 c:\Users\mhegazy\Documents\Visual Studio 2015\Projects\TypeScriptHTMLApp6\TypeScriptHTMLApp6\file1.tsx 10
Error TS17002 Expected corresponding JSX closing tag for 'div'. TypeScriptHTMLApp6 c:\Users\mhegazy\Documents\Visual Studio 2015\Projects\TypeScriptHTMLApp6\TypeScriptHTMLApp6\file1.tsx 10
Error TS1109 Expression expected. TypeScriptHTMLApp6 c:\Users\mhegazy\Documents\Visual Studio 2015\Projects\TypeScriptHTMLApp6\TypeScriptHTMLApp6\file1.tsx 10
Error TS2339 Property 'div' does not exist on type 'HTMLAttributes'. TypeScriptHTMLApp6 c:\Users\mhegazy\Documents\Visual Studio 2015\Projects\TypeScriptHTMLApp6\TypeScriptHTMLApp6\file1.tsx 10
Error TS2322 Type 'boolean' is not assignable to type 'number'. TypeScriptHTMLApp6 c:\Users\mhegazy\Documents\Visual Studio 2015\Projects\TypeScriptHTMLApp6\TypeScriptHTMLApp6\file1.tsx 10
```
| Bug,Help Wanted | low | Critical |
103,822,994 | three.js | BoundingSphere per BufferGeometry.group | At the moment `BufferGeometry`'s groups are sorted using the centre of the whole BufferGeometry. In the near future we could consider having a `BoundingSphere` per group so transparent sub parts of a geometry get proper sorted.
Follow up from #7066.
| Enhancement | low | Minor |
103,856,394 | rust | Local ambiguity in macro *definition* | ``` rust
macro_rules! m {
() => ();;
}
```
```
<anon>:2:14: 2:15 error: local ambiguity: multiple parsing options: built-in NTs tt ('lhs') or 1 other options.
<anon>:2 () => ();;
^
```
(the equivalent of `$( $lhs:tt => $rhs:tt );+ $(;)*` is used to parse macros)
This leaks "lhs", which is an implementation detail, and is a bad error message.
| A-diagnostics,P-low,A-macros,T-lang,C-bug | low | Critical |
103,859,130 | youtube-dl | Support for Discoverymax | Hi, I would like to ask you to add support for: www.discoverymax.marca.com.
The web offers the videos in a Brightcove playlist, and it works like this: There is a [playlist url](http://www.discoverymax.marca.com/series/otros/la-fiebre-del-oro/episodios-completos/#) and each video has it's _unique id_ ( i.e. #4383897233001).
If you type the _id_ at the end of the playlist you go to the video itself, but if you try downloading it with _youtube-dl_, you always get the first video of the playlist no matter what video's url you pass.
```
youtube-dl.exe -v http://www.discoverymax.marca.com/series/otros/la-fiebre-del-oro/episodios-completos/#4246353904001
[debug] System config: []
[debug] User config: []
[debug] Command-line args: [u'-v', u'http://www.discoverymax.marca.com/series/otros/la-fiebre-del-oro/episodios-completos/#4246353904001']
[debug] Encodings: locale cp1252, fs mbcs, out cp850, pref cp1252
[debug] youtube-dl version 2015.08.28
[debug] Python version 2.7.8 - Windows-7-6.1.7601-SP1
[debug] exe versions: ffmpeg N-71816-g02001ad, ffprobe N-71816-g02001ad
[debug] Proxy map: {}
[generic] #4246353904001: Requesting header
WARNING: Falling back on generic information extractor.
[generic] #4246353904001: Downloading webpage
[generic] #4246353904001: Extracting information
[generic] Brightcove video detected.
[download] Downloading playlist: La fiebre del oro | Discovery MAX
[generic] playlist La fiebre del oro | Discovery MAX: Collected 1 video ids (downloading 1 of them)
[download] Downloading video 1 of 1
[Brightcove] 4383781458001: Downloading webpage
[Brightcove] 4383781458001: Extracting information
[debug] Invoking downloader on u'http://discoveryint1.edgeboss.net/download/discoveryint1/byocdn/media/1378939881/201507/1387/1378939881_4383964414001_EHD-121135B-DMAX-SPAIN.mp4'
[download] Destination: La fiebre del oro - Cientos de onzas-4383781458001.mp4
[download] 100% of 563.58MiB in 04:42
```
Note that the requested video was 4246353904001 and the downloaded video was 4383781458001 instead.
Thank you.
| site-support-request,geo-restricted | low | Critical |
103,865,117 | go | gdb: provide a stack unwinder in Python | The recently released Gdb 7.10 added support for writing a frame unwinder in Python,
which is ideal for us.
doc: https://sourceware.org/gdb/current/onlinedocs/gdb/Unwinding-Frames-in-Python.html#Unwinding-Frames-in-Python
The issue tracks that.
| NeedsInvestigation | low | Major |
103,895,892 | go | cmd/compile: disambiguate package short names in "is not assignable" compile errors | I wrote some code that produced the following error in:
```
call.go:83: wrong type of argument 0 to Return for *mock_codesearch.MockCodesearcher.FindManifestsByNodeName: []codesearch.Result is not assignable to []codesearch.Result
```
I was confused because those two types are the same. Then I realized that on the left, codesearch referred to `google3/releasetools/rpdb/common/codesearch` while on the right it referred to `google3/social/automation/tools/actuators/codesearch`.
When the short package names are the same, the long name should be used in this "is not assignable" error message, as is done in other cases.
| compiler/runtime | low | Critical |
103,901,360 | youtube-dl | [req] e-learning udacity.com | Course page: https://www.udacity.com/course/viewer#!/c-ud032/l-698949179/m-762588556
| site-support-request | low | Major |
103,937,042 | go | encoding/xml: support QName values / expose namespace bindings | It's not uncommon for XML to contain QNames as element and attribute values, e.g.
```
<my-document xmlns:foo="http//..." >
<my-element>foo:bar</my-element>
</my-document>
```
In order to correctly unmarshal the value, you need to know the namespace bindings in effect for my-element, but Decoder doesn't appear to expose this information. A simple addition to encoding/xml of:
```
func (d *Decoder) NamespaceBindings() map[string]string {
return d.ns
}
```
allows unmarshallers to access the necessary information, for example, I can now write:
```
type QName struct {
Namespace string
Local string
}
func (qname *QName) UnmarshalXML(d *xml.Decoder, start xml.StartElement) error {
var s string
d.DecodeElement(&s, &start)
i := strings.Index(s, ":")
prefix := ""
if i >= 0 {
prefix = s[:i]
qname.Namespace = s[i+1:]
} else {
qname.Namespace = s
}
var ok bool
qname.Namespace, ok = d.NamespaceBindings()[prefix]
if !ok {
return errors.New("Unbound namespace prefix: " + prefix)
}
return nil
}
```
Arguably, something like the above, and a corresponding attribute unmarshaller could be provided on the standard xml.Name.
More discussion of this issue here:
https://groups.google.com/forum/#!searchin/golang-nuts/QName/golang-nuts/DexmVLQOJxk/whBaKK9ntHsJ
go version go1.5 darwin/amd64
| Thinking,NeedsFix,early-in-cycle | low | Critical |
104,090,902 | neovim | URI for terminal (term://...) | Current proposal:
```
hierarchical part
┌──────────────────┴──────────────────┐
terminal://username@localhost/usr/local/bin/bash:1234/path/cwd/?ls;echo foo;#optional-name
└───┬───┘ └───┬───┘ └───┬───┘└─────────┬────────┘└─┬─┘└───┬────┘└────┬─────┘└─────┬─────┘
scheme user host path to shell pid cwd query fragment
```
@justinmk
> > If you run term:ls or term:cmd it executes the command then leaves the shell
> > Perhaps that behavior should be changed.
I was thinking about this and I was wondering if there was a little overlap here with the `:term ls` command and the `:!ls` command since both execute the command and then leave the shell.
| needs:design,terminal | medium | Major |
104,099,188 | rust | TLS destructors on the main thread are not always run | There are some platforms where TLS destructors are run when the main thread exits, and there are some platforms where this does not happen,~~ and there are some platforms where things just go crazy ~~. For example, testing this program:
``` rust
struct Foo;
impl Drop for Foo {
fn drop(&mut self) {
println!("wut");
}
}
thread_local!(static FOO: Foo = Foo);
fn main() {
FOO.with(|_| {});
}
```
This should print `Foo dtor`, but prints nothing on some targets:
- [ ] musl (e.g. i686-unknown-linux-musl, x86_64-unknown-linux-musl)
- [ ] android (e.g. arm-linux-androideabi)
- [ ] Linux with [glibc before 2.18](https://github.com/rust-lang/rust/pull/126829#issuecomment-2187674002)
- [ ] Maybe more?
Here's a more complicated testcase that also involves initializing a destructor while destructors are being run; ideally this will be added to the test suite at some point:
```rust
struct Bar;
impl Drop for Bar {
fn drop(&mut self) {
println!("Bar dtor");
}
}
struct Foo;
impl Drop for Foo {
fn drop(&mut self) {
println!("Foo dtor");
// We initialize another thread-local inside the dtor, which is an interesting corner case.
thread_local!(static BAR: Bar = Bar);
BAR.with(|_| {});
}
}
thread_local!(static FOO: Foo = Foo);
fn main() {
FOO.with(|_| {});
}
```
### some older notes
- Linux ELF TLS - appears to work
- Linux pthread destructors - appear to not work (#19776)
- OSX - ~~appears to call destructors, but the program above specifically causes some form of memory corrupting, triggering an assert in malloc~~ fixed by https://github.com/rust-lang/rust/pull/57655
- Windows GNU/MSVC - appears to ~~not~~ work. ~~We're [listening for `DLL_PROCESS_DETACH`](https://github.com/rust-lang/rust/blob/6d6bf81b33abf3c8e200532b5f6f9e72ba540e92/src/libstd/sys/windows/thread_local.rs#L246) but for some reason we're not getting that notification.~~
| A-destructors,E-needs-test,T-libs-api,A-thread-locals,C-bug | medium | Major |
104,105,131 | opencv | the tutorial of high dynamic range imaging may cause misleading | http://docs.opencv.org/master/d3/db7/tutorial_hdr_imaging.html#gsc.tab=0
despite I didn't really try that,
the outcome of the tone mapped HDR should not be so poor, compared to the result of original paper
Their may be some bugs occur in the overall procedure
Durand, Frédo, and Julie Dorsey. "Fast bilateral filtering for the display of high-dynamic-range images." ACM transactions on graphics (TOG) 21.3 (2002): 257-266.
| bug,priority: low,affected: 3.4,category: photo | low | Critical |
104,143,409 | go | x/tools/cmd/goimports: treat import paths as standard library only if they are in goroot | The go tool sets a field to indicate whether a package is part of the standard library:
https://github.com/golang/go/blob/8d478e845c/src/cmd/go/pkg.go#L156
```
p.Standard = p.Goroot && p.ImportPath != "" && !strings.Contains(p.ImportPath, ".")
```
goimports uses a different rule:
https://github.com/golang/tools/blob/e21b7325f7/imports/fix.go#L32
```
if strings.Contains(importPath, ".") {
```
My feature request is to change the code to behave similar to the go tool:
```
if strings.Contains(importPath, ".") || !inGoroot(importPath) {
```
With an additional index of the standard library keyed by import path, I'd expect this to be reasonably fast.
If this is an acceptable idea, I'm happy to write a patch.
| Tools | low | Minor |
104,152,115 | youtube-dl | [req] e-learning -- sitepoint.com | Sitepoint has lots of video courses.
Course page: https://www.sitepoint.com/premium/courses/introduction-to-git-2902
| site-support-request,account-needed | low | Minor |
104,166,707 | TypeScript | Object.constructor isn't implemented specifically enough | Consider:
``` ts
class Foo extends ([].constructor) {
}
```
`[].constructor` should be `Array`, which we aught to be able to extend. However, the checker complains on `[].constructor` that "Type 'Function' is not a constructor function type". Considering `.constructor` is _quite explicitly_ the constructor function for the object, I believe this is a lie perpetuated by our `lib.d.ts` files. I think we need to go add appropriate self-constructor-referencing `.constructor` fields to each builtin.
_Additionally_ consider the following:
``` ts
class Bar {
}
class Foo extends ((new Bar()).constructor) {
}
```
The same thing but with a custom class rather than a builtin. We should know the correct type of `.constructor` here (the `Bar` class), but we don't - we just look at the members on `Object`.
In the first example, I believe it could be fixed with more precise lib files, but in the second case, I believe this is an instance of a field whose type should be known by the compiler, much like `.prototype` on the class itself is known.
| Bug,Help Wanted | low | Major |
104,171,157 | react | Play Nicely with The DOM Event System (because it's legacy anyway) | Currently React implements its own plugin system and event bubbling rules. This is important for systems like the responder system since the DOM event system is incapable of implementing proper event negotiation support.
However, on desktop, we don't really take full advantage of our custom event system other than to polyfill mouseenter/leave and implement more bubbling.
There are plenty of other things that are wrong with the DOM but instead of patching it, maybe it would be better to build a new view system on top of the DOM that integrates with the custom event system. Similarly to how React Native handles things.
That way, the lowest level could just do what the DOM would do, however broken it may be.
Question though: How do we handle things like onChange?
cc @spicyj and @syranide since I know you always wanted this.
| Type: Big Picture,React Core Team | low | Critical |
104,171,915 | go | x/net/websocket: Asymmetric treatment of Origin | In the docs for websocket.Handler it says you can ignore Origin, for non-browser requests as an example. However the Dial\* functions all throw errors if an empty origin is provided.
| NeedsInvestigation | low | Critical |
104,241,081 | opencv | BFMatcher doesn't work with MatVector | ```
fastDetector->detect(l, imgL_keypoints);
descriptor->compute(l, imgL_keypoints, imgL_descriptors);
vector<Mat> descriptorsL;
for (int i = 0; i < imgL_descriptors.rows; i++)
{
descriptorsL.push_back(imgL_descriptors.row(i));
}
fastDetector->detect(r, imgR_keypoints);
descriptor->compute(r, imgR_keypoints, imgR_descriptors);
vector<Mat> descriptorsR;
for (int i = 0; i < imgR_descriptors.rows; i++)
{
descriptorsR.push_back(imgR_descriptors.row(i));
}
vector<DMatch> matches;
BFMatcher matcher;
matcher.match(descriptorsL, descriptorsR, matches);
```
This code breaks on the last line.
Console writes out:
```
OpenCV Error: Assertion failed (masks.size() == imageCount) in
cv::DescriptorMatcher::checkMasks, file ..\..\..\modules\features2d\src\matchers.cpp, line 605
```
I read [here](http://docs.opencv.org/master/d4/d32/classcv_1_1__InputArray.html#gsc.tab=0) that OpenCV's `InputArray` can take `vector<Map>` so I assume this is a bug?
| bug,priority: normal,category: features2d | low | Critical |
104,390,241 | youtube-dl | Unable to download this ESPN's film vidoe. | Hi!
$ youtube-dl -v http://espn.go.com/30for30/film?page=the-schedule-makers
[debug] System config: []
[debug] User config: []
[debug] Command-line args: [u'-v', u'http://espn.go.com/30for30/film?page=the-schedule-makers']
[debug] Encodings: locale UTF-8, fs UTF-8, out UTF-8, pref UTF-8
[debug] youtube-dl version 2015.08.28
[debug] Python version 2.7.3 - Linux-3.2.0-4-amd64-x86_64-with-debian-7.9
[debug] exe versions: avconv 1.2.12, avprobe 1.2.12, ffmpeg 1.2.12, ffprobe 1.2.12, rtmpdump 2.4
[debug] Proxy map: {}
WARNING: The program functionality for this site has been marked as broken, and will probably not work.
[ESPN] film?page=the-schedule-makers: Downloading webpage
ERROR: Unable to extract video id; 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.
Traceback (most recent call last):
File "/home/ant/bin/youtube-dl/youtube_dl/YoutubeDL.py", line 655, in extract_info
ie_result = ie.extract(url)
File "/home/ant/bin/youtube-dl/youtube_dl/extractor/common.py", line 287, in extract
return self._real_extract(url)
File "/home/ant/bin/youtube-dl/youtube_dl/extractor/espn.py", line 45, in _real_extract
webpage, 'video id')
File "/home/ant/bin/youtube-dl/youtube_dl/extractor/common.py", line 575, in _search_regex
raise RegexNotFoundError('Unable to extract %s' % _name)
RegexNotFoundError: Unable to extract video id; 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.
Thank you in advance. :)
| bug | low | Critical |
104,396,865 | go | spec: clarify how unsafe.Pointers may be used to determine offsets between addresses | For background see discussion at https://groups.google.com/d/topic/golang-nuts/Uet5p_3JhZs/discussion
Currently it is possible to determine the offset between two values' addresses by finding the difference between uintptrs converted from unsafe.Pointer values. This is the only way to be able to determine whether two slices overlap in memory since the advent of three index slicing and the only way to determine in less than linear time whether blocked sparse slices overlap (relevant for views in constructed 2D slices for e.g. BLAS operations). In the event that a moving GC is implemented this may no longer be safe; if a GC move occurs between taking the conversion to uintptr of address of the first and second values, the offset may be incorrect. Note that the GC move does not otherwise invalidate the offset between overlapping slices or falsely result in non-overlapping being construed as overlapping.
| NeedsInvestigation | high | Critical |
104,401,091 | electron | Support two-finger swipe navigation | OSX supports various gestures with the trackpad. One of the most useful of these is the two-finger swipe which allows you to navigate forward and backward in web apps. This could be very useful for electron apps.
| enhancement :sparkles:,pr welcome | high | Critical |
104,422,381 | go | cmd/go: 'go test' should dynamically compute best timeout | Using `go1.5`
I discovered that "go test" crashes after the timeout of 600s has passed. I was running all the benchmarks in `compress/flate` and I set `-benchtime 10s`. Since each benchmark will run for about 10s or more, and there are 36 benchmarks in that package, this occupies at least 360s, which is running close to the default of 600s.
In situations like these, the `go test` tool should be able to figure out that a timeout of 600s is cutting things close when running 36 benchmarks each for 10s and choose a timeout that is more appropriate.
| NeedsDecision | low | Critical |
104,582,139 | youtube-dl | specify config file | there's an option to --ignore-config is there any plans to allow specify a specific config file, that is multiple config files.?
Examples, downloading certain file types from specific sites. Downloading with username and password and parameters depending which sites you're downloading from.
Ultimately, I guess it's no more cumbersome than having a shell script, with command line options for specific downloads and/or using --ignore-config as needed.
DJ
| request | low | Minor |
104,636,103 | go | x/mobile/exp/sensor: simplify os thread proxifying mechanism on Android | The Android implementation of the sensor package need to initialize a ALooper. ALoopers must be accessed from the same OS thread, therefore each operation on the package must be done on a goroutine that is locked to an OS thread.
The current mechanism accepts new messages through a signal channel depending on the signals type, invokes the required functions and sets the return values. It requires each function call to be represented with a different input signal.
This mechanism could be simplified by making the locked goroutine consuming a chan of functions.
``` go
type inOut func(done chan struct{})
m.inOut <- func(done chan struct{}) {
// do something on the locked goroutine
// when done, close done.
close(done)
}
```
| mobile | low | Minor |
104,659,720 | go | runtime: c-shared with LD_PRELOAD does not work for some programs | I'm trying to make my own LD_PRELOAD-able library like this:
``` go
// compile: go build -buildmode=c-shared -o example.so example.go
// tested with Go 1.5 on Ubuntu 15.04 amd64
package main
import "C"
func main() {}
//export dummy
func dummy() C.int { return 42 }
```
This `example.so` works well with several simple programs such as `ls`, `cat`, `cp` and so on.
```
$ LD_PRELOAD=./example.so ls
```
However, some programs such as `firefox`, `google-chrome`, `gdb` get segfault immediately (or sometimes just get hung up).
I suspect this is a bug of the Go runtime.
Anyone can please look on this?
Note that C-implementation seems working well with any program:
``` c
// compile: gcc -Wall -fPIC -shared -o example-c.so example.c
int dummy() { return 42; }
```
| compiler/runtime | low | Critical |
104,662,761 | TypeScript | Un used variable is emmitted while transpiling enums. | # TypeScript code
<pre>
<code>
module Workspace {
export enum State {
IN_PROGRESS,
BUFFER_RENDERED,
SKIPPED,
TIME_OUT
}
}
</code>
</pre>
# Compled JavaScript Code.
<pre>
<code>
var Workspace;
(function (Workspace) {
(function (State) {
State[State["IN_PROGRESS"] = 0] = "IN_PROGRESS";
State[State["BUFFER_RENDERED"] = 1] = "BUFFER_RENDERED";
State[State["SKIPPED"] = 2] = "SKIPPED";
State[State["TIME_OUT"] = 3] = "TIME_OUT";
})(Workspace.State || (Workspace.State = {}));
var State = Workspace.State; // this code is use no where.
})(Workspace || (Workspace = {}));
</code>
</pre>
I am using TypeScript version 1.5.3
| Suggestion,Help Wanted | low | Major |
104,676,371 | go | cmd/internal/obj: lacks tests | There are no tests that check the validity of the generated bytestream against some known golden sample.
| compiler/runtime | low | Minor |
104,721,101 | gin | Pass var to route handler | Hi,
how can I pass a var (db connection pointer) to a route handler like this:
apiV1.GET("/users/:userid", userHandler)
so that I can use in userHandler ?
Now I use a global var but I want to move the handler functions to another package
| question | medium | Major |
104,743,718 | TypeScript | static property inheritance complaining when it shouldn't | I've seen that you've already closed many issues on this. But I'm reporting another one in hopes to wear you down ;)
It seems I can't change the signature of a static function in a derived class. The fact that the subclass is at all aware of the static functions on the super class is really strange.
I developed in C# for years, so I know what you were _trying_ to do... but given that you don't support a `new` keyword like C# does, I think this behavior is really wrong. If you're looking to target being a superset of ES6/ES7, you need to correct this behavior ASAP. Generally in JavaScript, static properties are not copied to subclasses unless the copy is explicit.
I ran into this issue with [RxJS Next](https://github.com/ReactiveX/RxJS), where I Subject must inherit from Observable, but they both need _different_ static `create()` signatures. So I end up hacking around and fighting TypeScript.
| Suggestion,Committed | medium | Critical |
104,780,490 | go | x/build: submit queue, possibly implemented as optional auto-submit on +2 when trybots/slowbots with rebase pass | I see a trend of giving +2 together with trybot recently.
because we've defined that +2 means it's ok to submit,
I wonder if trybot could just submit the change (on behave
of the user who gives +2 and also invokes the trybot) if
the trybot runs are successful.
what do you think?
| Builders,FeatureRequest | medium | Major |
104,786,118 | go | x/mobile/event/mouse: include button state in event | The Button field records which button is transitioning, but for some applications (like chording) it would be nice if there was also a bitmask that said which buttons are up and down. This value can be integrated by following the transitions, but it's arguably something that could be done once by the mouse event driver.
| mobile | low | Minor |
104,786,264 | go | x/mobile/event/mouse: implement helper method for modifier keys | It is conventional that on a laptop with only a trackpad button, option and command (on Macs) turn a click into a middle or right click. It would be nice to have this functionality provided once, in the mouse package.
| mobile | low | Minor |
104,805,941 | youtube-dl | Site Support Request: Popular Science / Portico Player | ```
$ youtube-dl --version
2015.09.03
```
```
$ youtube-dl -g http://www.popsci.com/thorium-dream
WARNING: Falling back on generic information extractor.
http://www.popsci.com/\/\/www.popsci.com\/sites\/all\/modules\/bonnier\/bonnier_analytics\/providers\/omniture\/js\/s_code.js
```
The video is ultimately a straight-up MP4 inside a `video` tag, but it's injected via Javascript. Would love it if there were a way to support this kind of presentation.
| site-support-request | low | Minor |
104,875,071 | opencv | about pyopencv_to in cv2.cpp | Hey guys,
I am trying to compile OpenCV with ICC compiler. The code is from github , the latest version. When python component was compiled, it showed me this error message :
```
[ 64%] Building CXX object modules/python2/CMakeFiles/opencv_python2.dir/__/src2/cv2.cpp.obj
cv2.cpp
.\pyopencv_generated_types.h(1838): error: function "pyopencv_to(PyObject *, T &, const char *) [with T=cv::Moments]" was referenced but not defined
template<> bool pyopencv_to(PyObject* src, cv::Moments& dst, const char* name);
^
```
And then, I located the file `pyopencv_generated_types.h` and found out it was just a declaration
``` CPP
template<> bool pyopencv_to(PyObject* src, cv::Moments& dst, const char* name);
```
And I also found it in cv2.cpp:
``` CPP
template<typename T> static
bool pyopencv_to(PyObject* obj, T& p, const char* name = "<unknown>");
```
It that normal, or the issue of code generation ?
Gerron Jo
| bug,category: python bindings,priority: low,affected: 3.4 | low | Critical |
104,948,386 | go | net: allow custom Resolver method implementation(s) | I mentioned in #12476 that I would like to detect the time it took for DNS resolution phase of the Dial process in `Dialer.Dial` . The solution posted there was very hacky and adds unnecessarily to the API.
@bradfitz suggested
> Perhaps `net.Dialer` could have an optional `Resolver`, akin to how `http.Client` has an optional `Transport`, or `http.Server` has an optional `ErrorLog`, etc.
This seems like an excellent idea. Here is how I propose we go about it by adding minimal complexity and preserving code compatibility.
I propose net package adds a new `Resolver` interface.
```
type Resolver interface {
Resolve(host string, deadline time.Time) (addrs []IPAddr, err error)
}
```
The signature of `Resolver.Resolve` is same as `lookupIPDeadline` which `Dial` eventually uses. `Dialer` gets an optional field `CustomResolver` of type `Resolver`.
The `Resolver` object (or `nil`) gets passed around thru the resolution process.
`Dialer.Dial` -> `resolveAddrList` -> `internetAddrList` .
`internetAddrList` currently always uses `lookupIPDeadline`, it would need to be changed such that if the passed custom resolver is not nil then use it, otherwise use `lookupIPDeadline`.
Other functions calling `resolveAddrList` or `internetAddrList` would need to be modified to add an extra `nil` argument . This does not break code compatibility because they are unexported functions.
Benefits of allowing a custom `Resolver`
- Allowing me to collect timing information as mentioned in #12476
- Allowing users to implement their own DNS logic. Failovers, etc.
- Mocking for tests.
- Client side caching, pre-fetching, etc.
- In time, other packages ( like the superb github.com/miekg/dns ) could provide their own `Resolver` implementations.
- Great for people like me who rely on Go to write network debugging tools.
| help wanted,Proposal,Proposal-Accepted | high | Critical |
104,990,813 | go | x/build: bring back "broke the build" notifications | I monitor the dev.ssa dashboard for build failures, with decreasing frequency these days. It'd be nice to bring back the automatic notification.
cc @bradfitz
| help wanted,Builders,NeedsInvestigation,FeatureRequest | low | Critical |
105,024,637 | go | x/mobile/example/audio: build on windows system,run CRASH | Hello.
I am using gomobile in Windows System.
When build the example named "Audio",it is successful.Howevery,when run it on the Genymotion virtual device and a true Android device, the application is crashed the same.
After the debug, something I found:
- I can find the error message from logcat, it show this:
```
I/GoLog ( 9589): al: cannot load libopenal.so
```
- Additionally, I try to delete code line, then I am sure this code (in onStart function) throw this error:
```
rc, err := asset.Open("boing.wav")
```
Is someone get the same error when running the example of "Audio" ?
---
I am not sure is this the issues from gomobile,or other reason?
However, as a new developer who use gomobile to build android apps,I read the document from:
https://godoc.org/golang.org/x/mobile/example/audio
It tell me just to use "gomobile build".
Additionally, I notice the message about:
```
On Linux, you need to install OpenAL developer library by running the command below.
$ apt-get install libopenal-dev
```
However, on Windows, do I need to do something similarly?
If the GoDoc could write down it more clearly, we walked some detours can be avoided. :)
---
I am using the newest gomobile package and keep to update everyday.
So everything is newest when this problem was happended.
Thank you sincerely.
| OS-Windows,mobile | low | Critical |
105,043,756 | go | runtime: throw in linked c++ library causes app crash even if caught, on windows platform | when compiling go code using cgo to link a c++ application that uses try/catch/throw logic, the golang app crashes immediately when an exception is thrown. on linux, it performs as expected.
code versions:
go version go1.5 windows/amd64
tdm64-gcc-5.1.0-2
the code that demonstrates issue can be found in https://gist.github.com/superbaddude/b52e60f2e8dee0868af2
the output from command on windows:
```
go test
```
```
=== RUN Test_Callthrow_and_catch_exception
Exception 0x20474343 0x1045e80 0x6161616161616161 0x7fff56cda1c8
PC=0x7fff56cda1c8
signal arrived during external code execution
github.com/superbaddude/testcgo._Cfunc_throw_and_catch_exception()
github.com/superbaddude/testcgo/_test/_obj_test/_cgo_gotypes.go:37 +0x38
github.com/superbaddude/testcgo.Callthrow_and_catch_exception()
c:/Code/go/src/github.com/superbaddude/testcgo/throwe.go:9 +0x1b
github.com/superbaddude/testcgo.Test_Callthrow_and_catch_exception(0xc08208e000)
c:/Code/go/src/github.com/superbaddude/testcgo/throwe_test.go:6 +0x1b
testing.tRunner(0xc08208e000, 0x702d00)
c:/go/src/testing/testing.go:456 +0x9f
created by testing.RunTests
c:/go/src/testing/testing.go:561 +0x874
goroutine 1 [chan receive]:
testing.RunTests(0x5e4f28, 0x702d00, 0x1, 0x1, 0x1)
c:/go/src/testing/testing.go:562 +0x8b4
testing.(*M).Run(0xc082033ee8, 0xc0820583d0)
c:/go/src/testing/testing.go:494 +0x77
main.main()
github.com/superbaddude/testcgo/_test/_testmain.go:54 +0x11d
goroutine 17 [syscall, locked to thread]:
runtime.goexit()
c:/go/src/runtime/asm_amd64.s:1696 +0x1
rax 0x0
rbx 0x1045e80
rcx 0x0
rdi 0x1045ec0
rsi 0x6d6850
rbp 0x0
rsp 0x24fc00
r8 0x10
r9 0x8735d0
r10 0x1
r11 0x3
r12 0x8
r13 0x5e2c36
r14 0x9
r15 0x8
rip 0x7fff56cda1c8
rflags 0x206
cs 0x33
fs 0x53
gs 0x2b
exit status 2
FAIL github.com/superbaddude/testcgo 0.914s
```
| help wanted,OS-Windows,NeedsInvestigation | medium | Critical |
105,478,672 | angular | component events should mask event propagation. | ```
@Component({
selector: 'my-component'
events: ['change']
})
@View({
template: '<input (change)="...">'
})
class MyComponent {
}
```
The above code creates an issue. It conflates the `change` event from the `MyComponent` with the bubbling `change` event from the `<input>`.
The correct behavior is that the `change` event which is part of the components should filter out the DOM `change` events. Otherwise
```
<my-component (change)="doSomething($event)">
```
will sometimes get the `$event` from the `MyComponent` and sometimes from the `input` event.
| type: bug/fix,area: core,state: confirmed,core: event listeners,core: inputs / outputs,P3 | medium | Critical |
105,490,336 | TypeScript | NavigateTo inappropriately writes "module" instead of "namespace" in display | 
| Bug,Help Wanted | low | Minor |
105,591,273 | opencv | cv::BFMatcher::match crashes graphics driver | I found a problem in the GPU brute force matcher. When a certain amount of points/data is used, the program crashes the graphics driver with the following exception:
OpenCV Error: Assertion failed (clEnqueueReadBuffer(q, (cl_mem)u->handle, CL_TRU
E, 0, u->size, alignedPtr.getAlignedPtr(), 0, 0, 0) == CL_SUCCESS) in cv::ocl::O
penCLAllocator::map, file ......\opencv\sources\modules\core\src\ocl.cpp, line
4612
......\opencv\sources\modules\core\src\ocl.cpp:4612: error: (-215) clEnqueueRe
adBuffer(q, (cl_mem)u->handle, CL_TRUE, 0, u->size, alignedPtr.getAlignedPtr(),
0, 0, 0) == CL_SUCCESS in function cv::ocl::OpenCLAllocator::map
This is caused by cv::BFMatcher::match and occurs with both 3D points and descriptors of higher dimensions and might be related to issue #4876. I tested this on 4 different machines and 5 different graphics adapters (Geforce 260, 280, 480, 660 and 970, all with newest drivers). The only machine where this did not happen, was the one with the GeForce 480. The crash always occurs, if a certain amount of points is used. However, all cards have enough memory. The following code reproduces the problem with 3D points (adjust the number of points accordingly):
```
unsigned int numPointsA = 100000;
unsigned int numPointsB = numPointsA;
cv::Mat pointsA((int) numPointsA, 3, CV_32FC1);
cv::Mat pointsB((int) numPointsB, 3, CV_32FC1);
for(unsigned int i = 0; i < numPointsA; i++)
{
pointsA.ptr<float>(i)[0] = (float) (rand()%100);
pointsA.ptr<float>(i)[1] = (float) (rand()%100);
pointsA.ptr<float>(i)[2] = (float) (rand()%100);
}
for(unsigned int i = 0; i < numPointsB; i++)
{
pointsB.ptr<float>(i)[0] = (float) (rand()%100);
pointsB.ptr<float>(i)[1] = (float) (rand()%100);
pointsB.ptr<float>(i)[2] = (float) (rand()%100);
}
cv::BFMatcher bfMatcher(cv::NORM_L2, false);
vector<cv::DMatch> matches12;
cv::UMat samplesFixed1U = pointsB.getUMat(cv::ACCESS_READ);
cv::UMat pointsMoved1U = pointsA.getUMat(cv::ACCESS_READ);
bfMatcher.match(samplesFixed1U, pointsMoved1U, matches12);
```
| bug,priority: normal,affected: 3.4,category: ocl | low | Critical |
105,626,263 | You-Dont-Know-JS | "types & grammar": ch4, JSON Stringification is not always a string | Big fan, read all the books, have some remarks on "Behavior Delegation" but really, Great sequel!
In ch4, "JSON Stringification", you wrote:
> ... the serialization result is always a string
but that's not true
``` javascript
typeof JSON.stringify() === "undefined"
```
| for second edition | low | Minor |
105,643,062 | You-Dont-Know-JS | "types & grammar": ch4, a few more "crazy" comparisons | It may be worthwhile to point out a few more ["loose equals" (`==`) oddities](https://github.com/getify/You-Dont-Know-JS/blob/master/types%20&%20grammar/ch4.md#the-crazy-ones) resulting from Booleans being coerced to Numbers, and Objects to their primitive values. I'm thinking specifically of Arrays:
``` js
// These are all true:
console.log([] == false);
console.log([0] == false);
console.log(['0'] == false)
console.log([''] == false)
console.log([' '] == false)
console.log([null] == false);
console.log([undefined] == false);
// Though, these are false:
console.log([false] == false);
console.log([NaN] == false);
// These are both true:
console.log([1] == true);
console.log(['1'] == true)
// While these are false:
console.log([42] == true);
console.log(['42'] == true);
console.log([true] == true);
```
You may not consider that behavior worth noting, and if so I would not argue with you :)
| for second edition | medium | Minor |
105,658,333 | thefuck | changing package name when I use "sudo apt-get install" | I know there shouldn't be nothing wrong with the command `sudo apt-get install numix-wallpaper-aurora` itself, but I think thefuck shouldn't be suggesting me to install other package.
This is my last commands history:
| help wanted | low | Minor |
105,668,030 | youtube-dl | feature reuquest to support http://www.wildsafarilive.com/ | I'm recently getting hooked on watching this website about mostly lions, and one of the rangers records it on screen and uploads it to youtube, however, it takes too long and I'm trying to automate it
you can also watch same live show at http://www.wildearth.tv/cam/wildearth-safaris
so, support for either of http://www.wildearth.tv/cam/wildearth-safaris or http://www.wildsafarilive.com/ is appreciated
| site-support-request | low | Minor |
105,713,024 | go | encoding/gob: panic in reflect.Value.IsNil when encoding via pointer type alias | [Here](http://play.golang.org/p/e0q5srlbgj) is a program containing a type `Foo` that implements `gob.GobEncoder` in a silly way: by calling back into `gob.Encoder.Encode` using a type alias in order to avoid recursing. (The real world use for this is to zero a problematic field and then gob-encode the rest.)
It causes `gob.Encoder.Encode` to panic in `reflect.Value.IsNil`:
```
panic: reflect: call of reflect.Value.IsNil on struct Value [recovered]
panic: reflect: call of reflect.Value.IsNil on struct Value [recovered]
panic: reflect: call of reflect.Value.IsNil on struct Value
goroutine 1 [running]:
encoding/gob.catchError(0x1044c3ec, 0x1043a0dc)
/usr/local/go/src/encoding/gob/error.go:38 +0xe0
encoding/gob.catchError(0x1044c46c, 0x1043a13c)
/usr/local/go/src/encoding/gob/error.go:38 +0xe0
reflect.Value.IsNil(0x1ad3c0, 0x10438dd0, 0xd9, 0x1042b9c8, 0x1042b9cc, 0xb31c0)
/usr/local/go/src/reflect/value.go:971 +0x120
encoding/gob.encIndirect(0x1ad3c0, 0x10438dd0, 0xd9, 0x1, 0x0, 0x0, 0x0, 0x52d0)
/usr/local/go/src/encoding/gob/encode.go:152 +0x60
encoding/gob.(*Encoder).encodeSingle(0x1044c400, 0x1044c420, 0x10438e20, 0x1ad3c0, 0x10438dd0, 0xd9)
/usr/local/go/src/encoding/gob/encode.go:291 +0x200
encoding/gob.(*Encoder).encode(0x1044c400, 0x1044c420, 0x1ad3c0, 0x10438dd0, 0xd9, 0x10435440)
/usr/local/go/src/encoding/gob/encode.go:694 +0x280
encoding/gob.(*Encoder).EncodeValue(0x1044c400, 0x18cf80, 0x10438dd0, 0x16, 0x0, 0x0)
/usr/local/go/src/encoding/gob/encoder.go:247 +0x780
encoding/gob.(*Encoder).Encode(0x1044c400, 0x18cf80, 0x10438dd0, 0x52d0, 0x0, 0x0)
/usr/local/go/src/encoding/gob/encoder.go:174 +0x80
main.Foo.GobEncode(0x11, 0x52d0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x10438b60)
/tmp/sandbox454377553/main.go:23 +0x120
main.(*Foo).GobEncode(0x10438b60, 0x1aa4c0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0)
<autogenerated>:1 +0xc0
encoding/gob.(*Encoder).encodeGobEncoder(0x1044c380, 0x1044c3a0, 0x104350e0, 0x1aa4c0, 0x10438b60, 0x16)
/usr/local/go/src/encoding/gob/encode.go:469 +0x100
encoding/gob.gobEncodeOpFor.func1(0x104353e0, 0x10435360, 0x1aa4c0, 0x10438b60, 0x16, 0x1)
/usr/local/go/src/encoding/gob/encode.go:613 +0x260
encoding/gob.(*Encoder).encodeSingle(0x1044c380, 0x1044c3a0, 0x10438db0, 0x1aa4c0, 0x10438b60, 0x16)
/usr/local/go/src/encoding/gob/encode.go:294 +0x2a0
encoding/gob.(*Encoder).encode(0x1044c380, 0x1044c3a0, 0x1aa4c0, 0x10438b60, 0x16, 0x104350e0)
/usr/local/go/src/encoding/gob/encode.go:694 +0x280
encoding/gob.(*Encoder).EncodeValue(0x1044c380, 0x1aa4c0, 0x10438b60, 0x16, 0x0, 0x0)
/usr/local/go/src/encoding/gob/encoder.go:247 +0x780
encoding/gob.(*Encoder).Encode(0x1044c380, 0x1aa4c0, 0x10438b60, 0x1fc26d, 0x0, 0x0)
/usr/local/go/src/encoding/gob/encoder.go:174 +0x80
main.main()
/tmp/sandbox454377553/main.go:39 +0x120
```
If I'm reading the documentation correctly, this program should work as expected (printing `Result: 17`). But even if it's not intended to work, it probably shouldn't cause a panic.
Note that if you change it so that the alias looks like `type Alias Foo`, the program works whether you call `Encode` with `Alias(f)` or `(*Alias)(&f)`.
Go version:
```
go version devel +da7e9e4 Thu Sep 10 00:59:04 2015 +0000 darwin/amd64
```
| NeedsInvestigation | low | Critical |
105,764,385 | youtube-dl | [site-support-request] DeutschlandRadioKultur, Deutschlandradio, DR Wissen | Pages like http://www.deutschlandradiokultur.de/bestsellerautor-marc-elsberg-wie-realistisch-sind-ihre.970.de.html?dram:article_id=330629 have mp3 "play" urls (in this specific case: http://ondemand-mp3.dradio.de/file/dradio/2015/09/10/drk_20150910_0907_9054fc27.mp3 ), but are currently not detected by youtube-dl.
Would be nice to have these Urls fetchable, too.
| site-support-request | low | Minor |
105,865,271 | go | x/mobile/bind: support types from external packages | @crawshaw
| mobile | low | Minor |
105,866,809 | go | x/mobile/cmd/gomobile: gomobile bind returning io.Reader errors | I'm running into an error when running "gomobile bind --target=ios":
gomobile: type io.Reader not defined in package package goax ("github.com/garry415/goax")
unsupported, direct named type github.com/garry415/goax.KeyExchange: struct{IdentityPublic []byte "bencode:\"identity\""; Dh []byte "bencode:\"dh\""; Dh1 []byte "bencode:\"dh1\""}
I'm importing "io" and KeyExchange has only private fields. I don't think it should report this error.
Environment: Go 1.5, OS X Yosemite 10.10.4
| NeedsInvestigation,mobile | low | Critical |
106,024,468 | youtube-dl | cant download from qq | Hi
I tried:
```
youtube-dl http://v.qq.com/page/y/i/0/y01647bfni0.html
[generic] y01647bfni0: Requesting header
WARNING: Falling back on generic information extractor.
[generic] y01647bfni0: Downloading webpage
[generic] y01647bfni0: Extracting information
ERROR: Unsupported URL: http://v.qq.com/page/y/i/0/y01647bfni0.html
```
[Supported sites](http://rg3.github.io/youtube-dl/supportedsites.html) lists some qq sites as supported though.
| site-support-request | medium | Critical |
106,115,035 | youtube-dl | Support for video.thethao247.vn | Dear all,
Can you help for getting direct link from this site: mecloud.vn
For examlple:
embed.mecloud.vn/play/SLJtqLj3IO.
Regards,
SSD
| site-support-request | low | Minor |
106,177,806 | go | database/sql: Ability to get drivername from instance | I'd like to be able to get the driver name..
eg
``` go
Db, err = sql.Open("mysql", "root:root@/somestuff")
fmt.Printf("Driver= ", Db.Driver.Name) << Need...
```
Case Example.. I want to derive an sqlx instance from an existing connection
``` go
// Currently
func SetupDb( driver string, db *sql.DB){
Dbx = sqlx.NewDb(db, driver)
}
// Wanted
func SetupDb( db *sql.DB){
Dbx = sqlx.NewDb(db, db.Driver.Name)
}
```
https://github.com/daffodil/go-mailkit/blob/master/postfixadmin/setup.go#L15
| FeatureRequest | medium | Critical |
106,179,722 | youtube-dl | can not combine --playlist-reverse and --playlist-start? | Hello everyone
i was playing with the --playlist-reverse and the --playlist start NUMBER options to try to download in chronological order and resume where i stopped (like the next day continue at the 26th video) and found that these two options do not work together.
im not sure if this is a bug or if they are not supposed to be combined at all (maybe the help should say so then): in this case i would like to make a feature request for it!
some examples to show what i mean:
1: playlist-start 1 - no reverse
C:\youtubedl>youtube-dl.exe -e --max-downloads 3 --playlist-start 1 https://www.youtube.com/user/youtube
Introducing YouTube Gaming
# ProudToLove - Celebrating Marriage Equality and LGBT Pride Month
The A-Z of YouTube: Celebrating 10 Years
2: playlist-start 3 - no reverse (works as expected #1 video now = #3 of the previous command)
C:\youtubedl>youtube-dl.exe -e --max-downloads 3 --playlist-start 3 https://www.youtube.com/user/youtube
The A-Z of YouTube: Celebrating 10 Years
# DearMe - What Advice Would You Give Your Younger Self?
Highlights From the YouTube Interview with President Obama
3: playlist-start 1 + reverse
C:\youtubedl>youtube-dl.exe -e --max-downloads 3 --playlist-reverse --playlist-start 1 https://www.youtube.com/user/youtube
YouTube on the tube!
Hammer Time!
A Message From Chad and Steve
and finally
4: playlist-start 3 + reverse (BUG HERE: this gives the same result as command #3 above)
C:\youtubedl>youtube-dl.exe -e --max-downloads 3 --playlist-reverse --playlist-start 3 https://www.youtube.com/user/youtube
YouTube on the tube!
Hammer Time!
A Message From Chad and Steve
thanks in advance for feedback!
| request | medium | Critical |
106,197,666 | youtube-dl | List human-readable video title where possible in format list (-F) display | I often use `-F` to list the possible video encoding formats I can use for videos.
When I do this, I see something like
```
[info] Available formats for dQw4w9WgXcQ:
...
```
which tells me nothing about the video.
I've recently discovered that the formats available for YouTube videos is unpredictable and depends on what batch-transcoding settings were in place when the video was originally uploaded.
If I do an `-F` request for several videos, I'll get a terminal full of "Available formats for ....." and I'll then need to repeatedly crossreference the video IDs manually to work out what to do next - I might want to find a higher quality copy of a given video, or I might only want the audio track of another, etc, but I can only make the appropriate judgement calls if I know the title of the video in question!
I would suggest [re]using the filename output format specifier (minus the file extension, if feasible) to control the presentation.
Thanks for such an awesome program, btw! :D
| request | low | Minor |
106,320,712 | go | encoding/xml: XML CDATA section could be joined together with regular characters | go version go1.5 darwin/amd64
One thing I stumbled across yesterday (not a real bug, but a minor nuisance from a user's perspective perhaps):
```
package main
import (
"encoding/xml"
"fmt"
"strings"
)
func main() {
src := `<root>a<![CDATA[b]]>c</root>`
r := strings.NewReader(src)
dec := xml.NewDecoder(r)
for {
tok, err := dec.Token()
if err != nil {
fmt.Println(err)
break
}
fmt.Printf("%#v\n", tok)
}
}
```
gives
```
xml.StartElement{Name:xml.Name{Space:"", Local:"root"}, Attr:[]xml.Attr{}}
xml.CharData{0x61}
xml.CharData{0x62}
xml.CharData{0x63}
xml.EndElement{Name:xml.Name{Space:"", Local:"root"}}
EOF
```
I would expect one `xml.CharData{}` token instead:
```
xml.StartElement{Name:xml.Name{Space:"", Local:"root"}, Attr:[]xml.Attr{}}
xml.CharData{0x61, 0x62, 0x63}
xml.EndElement{Name:xml.Name{Space:"", Local:"root"}}
EOF
```
While I understand the source of the three tokens, I would expect one as the user (= me) is unable to distinguish between a CDATA node and a regular text node.
| NeedsInvestigation | low | Critical |
106,342,563 | opencv | Optimizing cvReprojectImageTo3D() | As part of an API release, I have published a method for projecting a disparity map to 3D points, similar to cvReprojectImageTo3D(). Unlike the OpenCV implementation, my approach makes use of SSE2 or AVX2 to gain significant speed-ups. The code has been published under the MIT license, which should allow a re-distribution under the BSD license (and if not, you have my permissions).
Code for the method in question can viewed here:
http://nerian.com/products/sp1-stereo-vision/api-doc/reconstruct3d_8cpp_source.html
The entire API source code can also be downloaded from this location:
http://nerian.com/products/sp1-stereo-vision/downloads/sp1-software-1_0_1-src.tar.xz
I think it would be nice to have these optimizations ported to OpenCV. However, there are some significant differences to the OpenCV method:
- Points in the produced point cloud are aligned to 16-byte boundaries, introducing 4 bytes of padding. But this could be changed using unaligned stores, with some performance penalties
- The method only works for 16-bit unsigned disparity maps. OpenCV only works with 8-bit unsigned, 16-bit signed, 32-bit signed or 32-bit floating.
- Invalid disparities are assumed to have a large value (>=0xFFF). OpenCV expects a small value.
- It is possible to specify a minimum disparity value
- All computations are performed with floats, rather than doubles.
So, I'm not sure if its worth porting that to OpenCV. I guess at at least the 16-bit signed case could be covered, but the handling of invalid disparities is rather different.
Any suggestions?
Konstantin
| priority: normal,feature,category: calib3d,affected: 3.4 | low | Major |
106,416,585 | youtube-dl | request: add option for re-download video in better quality | YouTube public video is not in the best quality to make it available as soon as possible (for video encoding in a better quality requires more time). If the video is downloaded in any capacity, the programa miss it. I would like to be able to re-download the video in better quality, instead of the old, if possible. Important! Previous video should not be removed as long as the new will not be fully loaded.
| request | low | Minor |
106,429,297 | neovim | default mapping: <c-r>? to put prettified search text | Hi,
when paste last search (with `<C-r>/` for insert mode or `"/p` for normal) : always remove `\< \>` has seen in <my_search>
scenario :
given
My current search is : "<foobar>"
when
In normal mode I press : "/p
then
It prints "foobar"
Currently it works with vimscript with that:
``` vimscript
" remove ugly char when pasted searched text
function! Del_word_delims()
let reg = getreg('/')
" After * i^r/ will give me pattern instead of \<pattern\>
let res = substitute(reg, '^\\<\(.*\)\\>$', '\1', '' )
if res != reg
return res
endif
" After * on a selection i^r/ will give me pattern instead of \Vpattern
let res = substitute(reg, '^\\V' , '' , '' )
let res = substitute(res, '\\\\' , '\\', 'g')
let res = substitute(res, '\\n' , '\n', 'g')
return res
endfunction
inoremap <silent> <C-R>/ <C-R>=Del_word_delims()<CR>
cnoremap <C-R>/ <C-R>=Del_word_delims()<CR>
nnoremap "/p "=Del_word_delims()<C-M>p
```
| enhancement,defaults | low | Minor |
Subsets and Splits