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 |
|---|---|---|---|---|---|---|
545,878,194 | vscode | Provide a way to debug extensions to hint syntax for advanced breakpoints | Right now, when you try to create a hit count breakpoint, you're given this:
And log points like so:
The data in here is text and can be interpreted however the debug adapter sees fit. It might be useful to allow the adapter to hint at what their syntax is to help the user figure out what to put in their without having to google for whatever adapter they're using.
Ideas:
- A default always-true placeholder. For JS debug, that would be the string `>= 0`. This would only allow showcasing a single hint, however, and would be less useful for logpoint breaks.
- A way for the adapter to provide a URL to find more information about the syntax they use, and surface that inline somehow.
- A message/markdown to be shown around the breakpoint UI. This might get too noisy with long messages, it could be hidden by default and toggled with a button.
cc @misolori
| debug,under-discussion | low | Critical |
545,909,675 | godot | Colored emojis' outlines are broken when defined as a fallback for DynamicFont | **Godot version:**
24281cd
**Issue description:**
If a `DynamicFont` with `outline` bigger than 0 has colored emojis defined as a fallback font, their outlines will be drawn incorrectly, likely drawing the outlines of the broken characters of the main.
**Minimal reproduction project:**
[example.zip](https://github.com/godotengine/godot/files/5699789/example.zip) | bug,confirmed,topic:gui | low | Critical |
545,944,422 | go | all: simplify standard library examples | Over the last few years, Go standard library has done a great job by improving the examples coverage. However, some of the examples are difficult to read or too comprehensive to be a reference for basic usage.
Standard library examples can help more if they were simpler and more copy/paste-able. More comprehensive scenarios can be covered as secondary examples or in other mediums such as https://gobyexample.com/.
Here are some examples:
- https://golang.org/pkg/path/filepath/#Walk
It creates a directory, adds a lot of logic, prints out unix-specific comments, etc. Instead, it can demonstrate how to walk in a well known path, such as the home directory.
- https://golang.org/pkg/io/#example_TeeReader
It doesn't really demonstrate a case where a TeeReader is useful. Running is required to understand what it does.
- https://golang.org/pkg/os/exec/#example_Cmd_StderrPipe
It is hard to follow what the sh is supposed to do with "echo stdout; echo 1>&2 stderr". It would be more valuable if the command is something everyone can relate.
Sorry for having to nitpick these examples, there are more cases in the standard library and it would be great to turn some of these examples into more digest ones.
Having said that, I see some packages (such as the io ones) might be better with comprehensive examples where edge cases are explained and demonstrable by running the example program because APIs are not always doing a great job explaining the behavior. | Documentation,help wanted,NeedsInvestigation | low | Major |
545,944,868 | pytorch | torch.masked_select out argument can easily be misused, because output shape is dynamically computed | ## 🐛 Bug
<!-- A clear and concise description of what the bug is. -->
Unexpected behavior of `masked_select`, likely due to `out` option should be contiguous but was not clarified in the doc.
## To Reproduce
Steps to reproduce the behavior:
```python
x, m, z = torch.ones((4,3,2)), torch.zeros((4,3), dtype=torch.bool), torch.zeros((4,2,2))
m[:, 1] = True
torch.masked_select(x, m[..., None], out=z[:, 1, :])
```
## Expected behavior
```python
>>> print(z)
tensor([[[0., 0.],
[1., 1.]],
[[0., 0.],
[1., 1.]],
[[0., 0.],
[1., 1.]],
[[0., 0.],
[1., 1.]]])
```
## Observed behavior
```python
>>> print(z)
tensor([[[0., 0.],
[1., 1.]],
[[1., 1.],
[1., 1.]],
[[1., 1.],
[0., 0.]],
[[0., 0.],
[0., 0.]]])
```
## Environment
PyTorch version: 1.3.1
CUDA used to build PyTorch: 10.1.243
OS: Ubuntu 18.04
Python version: 3.7
cc @ezyang @gchanan @zou3519 @mruberry @rgommers @heitorschueroff | high priority,triaged,module: numpy,module: safe resize | low | Critical |
545,950,893 | rust | Expose `abort` function from core | It seems like `core` knows how to abort a program (it has `core::intrinsics::abort`), but only `std` exposes this functionality.
In some rare cases, a potentially unwinding panic might lead to unsoundness, where an abort would do a better job. A good example is [the implementation of `Arc`](https://github.com/rust-lang/rust/blob/ebbb2bf37aedaaa64dfaa52ba337ca6efb6b9093/src/liballoc/sync.rs#L947-L960).
The very same situation occurs [in the `atomic_refcell` crate](https://github.com/bholley/atomic_refcell/blob/807df0a411b698412c4bc890649679f0aaa3ef20/src/lib.rs#L164-L173). As there is nothing else in the crate depending on `std` it should be a `no_std` crate. However, that change replaces the `std` exit/abort call with [an ugly hack using a double-panic](https://github.com/bholley/atomic_refcell/blob/195db87e0feaef15539e195e1c9cbc472e22a72e/src/lib.rs#L165-L188) to try to cause an abort.
Instead, a stable `abort` function in `core` would be a lot nicer. | T-lang,T-libs-api,C-feature-request,WG-embedded | low | Major |
545,981,567 | TypeScript | Weird error message for mixed exported/non-exported declarations under `isolatedModules` | ```ts
// @isolatedModules: true
export function bar() {
}
interface bar {
}
export { bar as MyThing };
```
**Expected**: I actually don't know 😄. I think it should just work though.
**Actual**: Unclear error message
In nightlies (3.8+), the message is currently
```
Re-exporting a type when the '--isolatedModules' flag is provided requires using 'export type'.
```
In 3.7.4 the message is still unclear.
```
Cannot re-export a type when the '--isolatedModules' flag is provided.
```
But `bar` is also a value!
See also
* https://github.com/microsoft/TypeScript/issues/32662 (which is about correcting the phrasing of the error)
* https://github.com/microsoft/TypeScript/issues/32601 (which is about I guess giving the user better suggestions) | Suggestion,Needs Proposal | low | Critical |
545,983,573 | node | rmdirsync doesn't throw error when symlinked path is passed as argument in windows node 12 | I have written a simple code to pin down the issue.
Env: Windows
Node version: 12
```js
const fs = require('fs');
fs.mkdirSync("tmp");
fs.symlinkSync(".", "tmp/out");
console.log(`fs.existSync(tmp/out) : ${fs.existsSync('tmp/out')}`);
fs.rmdirSync('tmp/out');
```
Execution in CI:
Node 12
```sh
Build started
git clone -q --branch=rmdirsync https://github.com/SparshithNR/node-symlink.git C:\projects\node-symlink
git checkout -qf a6cb35fc2b3d2df328d4aeae3db65d732243142c
Restoring build cache
Cache 'C:\Users\appveyor\AppData\Local\Yarn' - Restored
Running Install scripts
Install-Product node $env:nodejs_version
Uninstalling node 8.16.2 (x86)...
Installing node 12.13.1 (x86)...
git rev-parse HEAD
a6cb35fc2b3d2df328d4aeae3db65d732243142c
yarn run test
yarn run v1.16.0
$ node index.js
fs.existSync(tmp/out) : true
Done in 0.27s.
Updating build cache...
Cache 'C:\Users\appveyor\AppData\Local\Yarn' - Up to date
Build success
```
Node 10:
```sh
Build started
git clone -q --branch=rmdirsync https://github.com/SparshithNR/node-symlink.git C:\projects\node-symlink
git checkout -qf a6cb35fc2b3d2df328d4aeae3db65d732243142c
Restoring build cache
Cache 'C:\Users\appveyor\AppData\Local\Yarn' - Restored
Running Install scripts
Install-Product node $env:nodejs_version
Uninstalling node 8.16.2 (x86)...
Installing node 10.17.0 (x86)...
git rev-parse HEAD
a6cb35fc2b3d2df328d4aeae3db65d732243142c
yarn run test
yarn run v1.16.0
$ node index.js
fs.existSync(tmp/out) : true
fs.js:114
throw err;
^
Error: ENOENT: no such file or directory, rmdir 'tmp/out'
at Object.rmdirSync (fs.js:684:3)
at Object.<anonymous> (C:\projects\node-symlink\index.js:6:4)
at Module._compile (internal/modules/cjs/loader.js:778:30)
at Object.Module._extensions..js (internal/modules/cjs/loader.js:789:10)
at Module.load (internal/modules/cjs/loader.js:653:32)
at tryModuleLoad (internal/modules/cjs/loader.js:593:12)
at Function.Module._load (internal/modules/cjs/loader.js:585:3)
at Function.Module.runMain (internal/modules/cjs/loader.js:831:12)
at startup (internal/bootstrap/node.js:283:19)
at bootstrapNodeJSCore (internal/bootstrap/node.js:623:3)
error Command failed with exit code 1.
info Visit https://yarnpkg.com/en/docs/cli/run for documentation about this command.
Command exited with code 1
```
Code can be found here: https://github.com/SparshithNR/node-symlink/tree/rmdirsync
Windows Execution: https://ci.appveyor.com/project/SparshithNR/node-symlink/builds/29933274
Linux Execution: https://github.com/SparshithNR/node-symlink/runs/376483962 | fs | low | Critical |
546,005,120 | node | Provide built-in implementations of module linker, initializeImportMeta, importModuleDynamically | **Is your feature request related to a problem? Please describe.**
When using vm.SourceTextModule we have to implement a linker, initializeImportMeta, and importModuleDynamically, even if we just want the default behavior of built-in module support. There are many subtle ways to get that wrong.
**Describe the solution you'd like**
Default implementations that do basically what built-in module support does:
- A linker function that:
- Reads files from disk, within a specified root.
- Uses Node module resolution as implemented for modules
- Uses a specified module cache
- `initializeImportMeta` that sets `url` property
- By configuration, allows access to built-in modules.
**Describe alternatives you've considered**
Userland libraries can take the first cuts at easy-to-use vm.Modules, but I think something for common use cases should likely be included.
| vm | low | Major |
546,008,976 | rust | Mutable borrow of field with autoref vs ref mut | I've been playing around with a singly-linked-list. It turns out that this method, which I had hoped might compile with help of non-lexical-lifetimes, does not compile:
struct Recursive {
child: Option<Box<Recursive>>
}
impl Recursive {
fn mut_back(&mut self) -> &mut Self {
let mut node = self;
while let Some(child) = node.child.as_mut() {
node = child
}
node
}
}
The error is stating that the borrow for `node.child` overlaps with the returned borrow of `node`, as seen below:
error[E0499]: cannot borrow `*node` as mutable more than once at a time
--> src/lib.rs:11:13
|
6 | fn mut_back(&mut self) -> &mut Self {
| - let's call the lifetime of this reference `'1`
7 | let mut node = self;
8 | while let Some(child) = node.child.as_mut() {
| ---------- first mutable borrow occurs here
...
11 | node
| ^^^^
| |
| second mutable borrow occurs here
| returning this value requires that `node.child` is borrowed for `'1`
error: aborting due to previous error
There is a _very similar_ implementation of `mut_back()` which does compile:
impl Recursive {
fn mut_back_good(&mut self) -> &mut Self {
let mut node = self;
while let Some(ref mut child) = node.child {
node = child
}
node
}
}
What I think is going on to cause the first implementation to fail is the call to `child.as_mut()`. Presumably due to autoref this creates a new unique borrow on `node.child`.
In the second case, I understand that the `ref mut` binding means we're _also_ creating a unique borrow on `node.child`. But in this case the borrow checker doesn't think this new unique borrow is a problem.
It seems to me that the compiler should be able to treat these two implementations equally? | A-lifetimes,T-compiler,A-NLL,NLL-complete,NLL-polonius | low | Critical |
546,014,369 | vscode | Copying text from a Markdown document inserts non-breaking spaces instead of regular spaces |
Issue Type: <b>Bug</b>
1. Install VSCode on an entirely new machine that has never had it before (this is to show that extensions and settings have no impact on this behaviour).
2. Create a new Markdown document with several words in it, separated by regular spaces.
3. Copy the contents of the Markdown document to the clipboard. Use the editor to copy, not the preview window.
4. Paste into an application that highlights non-breaking spaces, such as Word. **Observe that every space has been replaced by a non-breaking space. This is unexpected behaviour.**
5. Change the format of the document in VSCode to 'Plain Text'. Copy the text again and paste back into Word. Observe that the spaces are copied out as regular spaces, which is the correct behaviour.
VS Code version: Code - Insiders 1.42.0-insider (9349e9a2fdbd1d92824ce99af4a50b6785351b11, 2020-01-06T10:40:10.218Z)
OS version: Windows_NT x64 10.0.18363
<details>
<summary>System Info</summary>
|Item|Value|
|---|---|
|CPUs|Intel(R) Core(TM) i7-8650U CPU @ 1.90GHz (8 x 2112)|
|GPU Status|2d_canvas: enabled<br>flash_3d: enabled<br>flash_stage3d: enabled<br>flash_stage3d_baseline: enabled<br>gpu_compositing: enabled<br>metal: disabled_off<br>multiple_raster_threads: enabled_on<br>oop_rasterization: disabled_off<br>protected_video_decode: enabled<br>rasterization: enabled<br>skia_renderer: disabled_off<br>surface_control: disabled_off<br>surface_synchronization: enabled_on<br>video_decode: enabled<br>viz_display_compositor: enabled_on<br>viz_hit_test_surface_layer: disabled_off<br>webgl: enabled<br>webgl2: enabled|
|Load (avg)|undefined|
|Memory (System)|15.93GB (7.62GB free)|
|Process Argv||
|Screen Reader|no|
|VM|0%|
</details>
<!-- generated by issue reporter --> | windows,under-discussion,editor-clipboard | low | Critical |
546,031,879 | TypeScript | JSDoc property name is not parsed as expected when non-alphanumeric characters are present and intellisense not showing properties | *TS Template added by @mjbvz*
**TypeScript Version**: 3.8.0-dev.20200104
**Search terms**:
- jsdoc / jsdocs
- completionInfo
---
<!-- Please read our Rules of Conduct: https://opensource.microsoft.com/codeofconduct/ -->
<!-- Please search existing issues to avoid creating duplicates. -->
<!-- Also please test using the latest insiders build to make sure your issue has not already been fixed: https://code.visualstudio.com/insiders/ -->
<!-- Use Help > Report Issue to prefill these. -->
- VSCode Version: 1.41.1 (user setup)
- OS Version: Windows_NT x64 10.0.17134
**Origin of this discovery:**
I was trying to document my types which translate to XML classes. The syntax for which includes the "@class" as a property name of my object. This is not possible given the current syntax parsing in VSCode.
**Steps to Reproduce:**
1. Place this code in the editor with the language as Javascript
Code:
```javascript
/**
* @typedef Foo
* @property {string} bar Property shows in intellisense
* @property {string} @baz Property should show in intellisense but does not
* @property {string} test Property should show in intellisense but does not
* @property {string} @class Property name "@class" should be a possible property name that
* does not get interpreted as a JSDoc instruction.
* @property {string} even#ThisShouldBeAValidName! But the name is not parsed as expeted
*/
/**
* @type {Foo}
*/
const fooInstance = {
}
```
2. Prompt the intellisense to show within the "fooInstance" object.
1. Expected: The properties @ baz and @ class should be highlighted like normal property names
1. Actual: The properties are shown as if they were JSDoc instructions
2. Expected: The property "even#ThisShouldBeAValidName!" should be highlighted like a normal property name
1. Actual: The property is highlighted as expected until the "#" symbol and then it seems to be treated like a comment
3. Expected: Intellisense should show all the properties
1. Actual: Only the first property "bar" is shown. An empty string "" is shown as if it were another property name.
<!-- Launch with `code --disable-extensions` to check. -->
Does this issue occur when all extensions are disabled?: Yes
| Suggestion,Needs Proposal | low | Minor |
546,036,777 | pytorch | Compile libtorch by source code failed. | Hi,guys, i am a newer for pytorch and libtorch, and i want to compile the libtorch by source official code,but there are some problem happened.I compile the source official code by this(https://github.com/pytorch/pytorch/blob/master/docs/libtorch.rst) guide directly(i don't compile the pytorch),but the problem happened like this:
[2066/2066] Install the project...
-- Install configuration: "Release"
CMake Error at third_party/protobuf/cmake/cmake_install.cmake:48 (file):
file INSTALL cannot find
"/home/cyj/Documents/pytorch-test/build_libtorch/build/third_party/protobuf/cmake/CMakeFiles/CMakeRelink.dir/protoc".
Call Stack (most recent call first):
cmake_install.cmake:80 (include)
FAILED: cd /home/cyj/Documents/pytorch-test/build_libtorch/build && /usr/bin/cmake -P cmake_install.cmake
ninja: build stopped: subcommand failed.
Traceback (most recent call last):
File "../tools/build_libtorch.py", line 23, in <module>
rerun_cmake=True, cmake_only=False, cmake=CMake())
File "/home/cyj/Documents/pytorch-test/tools/build_pytorch_libs.py", line 62, in build_caffe2
cmake.build(my_env)
File "/home/cyj/Documents/pytorch-test/tools/setup_helpers/cmake.py", line 339, in build
self.run(build_args, my_env)
File "/home/cyj/Documents/pytorch-test/tools/setup_helpers/cmake.py", line 141, in run
check_call(command, cwd=self.build_dir, env=env)
File "/usr/lib/python3.5/subprocess.py", line 581, in check_call
raise CalledProcessError(retcode, cmd)
subprocess.CalledProcessError: Command '['cmake', '--build', '.', '--target', 'install', '--config', 'Release', '--', '-j', '8']' returned non-zero exit status 1
I want to know the reason why i met such problem,Does any deploy i have not set? I just follow the official guide!My environment is ubuntu16.04 cuda10.0 gcc 5.4.0 cmake 3.5.1,the source code is newest.Hope for help,tkx.
cc @ezyang @yf225 | module: binaries,module: cpp,triaged | low | Critical |
546,053,528 | go | net: udp listener should use received ip address for sending | <!--
Please answer these questions before submitting your issue. Thanks!
For questions please use one of our forums: https://github.com/golang/go/wiki/Questions
-->
### What version of Go are you using (`go version`)?
<pre>
$ go version
go version go1.12.1 linux/amd64
</pre>
### Does this issue reproduce with the latest release?
Yes
### What operating system and processor architecture are you using (`go env`)?
<details><summary><code>go env</code> Output</summary><br><pre>
$ go env
GOARCH="amd64"
GOBIN=""
GOCACHE="/root/.cache/go-build"
GOEXE=""
GOFLAGS=""
GOHOSTARCH="amd64"
GOHOSTOS="linux"
GOOS="linux"
GOPATH="/data/go_work"
GOPROXY=""
GORACE=""
GOROOT="/usr/local/go"
GOTMPDIR=""
GOTOOLDIR="/usr/local/go/pkg/tool/linux_amd64"
GCCGO="gccgo"
CC="gcc"
CXX="g++"
CGO_ENABLED="1"
GOMOD=""
CGO_CFLAGS="-g -O2"
CGO_CPPFLAGS=""
CGO_CXXFLAGS="-g -O2"
CGO_FFLAGS="-g -O2"
CGO_LDFLAGS="-g -O2"
PKG_CONFIG="pkg-config"
GOGCCFLAGS="-fPIC -m64 -pthread -fmessage-length=0 -fdebug-prefix-map=/tmp/go-build134300587=/tmp/go-build -gno-record-gcc-switches"
</pre></details>
### What did you do?
<!--
If possible, provide a recipe for reproducing the error.
A complete runnable program is good.
A link on play.golang.org is best.
-->
At the very first time, we found that when I dig the backend dns server, which is written in go, we got an unexpect source alert, it seems that the reply came from another interface from the machine.
So I want to stimulate this issue with a tiny program like:
```go
package main
import (
"fmt"
"net"
)
func main() {
listener, err := net.ListenUDP("udp", &net.UDPAddr{IP: net.ParseIP("0.0.0.0"), Port: 9999})
if err != nil {
fmt.Println(err)
return
}
fmt.Printf("Local: <%s> \n", listener.LocalAddr().String())
data := make([]byte, 1024)
for {
n, remoteAddr, err := listener.ReadFromUDP(data)
if err != nil {
fmt.Printf("error during read: %s", err)
}
fmt.Printf("<%s> %s\n", remoteAddr, data[:n])
_, err = listener.WriteToUDP([]byte("world"), remoteAddr)
if err != nil {
fmt.Printf(err.Error())
}
}
}
```
UDP server listen on (:9999 or 0.0.0.0:9999)
### What did you expect to see?
The machine we are gonna access is bound with multiple interface, i.e. it has eth0/eth1/tun0/tun1 etc.
so when the client get access to this service from eth0, it is expect that the reply must be sent from interface eth0.
### What did you see instead?
Udp server local ip (multi)
x.x.52.207
x.x.56.205
x.x.78.207
client (remote IP) : x.x.218.174
run tcpdump from udp server
as you can see, the server always reply from the default interface, the steps are as belows:
1. client(x.x.218.174) -> udp server(x.x.52.207) got reply from udp server(x.x.52.207)
2. client(x.x.218.174) -> udp server(x.x.56.205) got reply from udp server(x.x.52.207)
TCP connections all work as expect except UDP connections.
However, when we test this scenario with Nginx, it works:
As you can see it reply the client with the incoming interface instead of the default one.
Here is the simplified config:
```
# nginx listen on 0.0.0.0:9999
stream {
upstream dns {
server x.x.41.44:9999;
}
server {
listen 9999 udp;
... ...
}
}
```
| NeedsInvestigation | low | Critical |
546,083,478 | PowerToys | Quick Display adjustment PowerToy | End users are asking for a tool to quickly toggle settings of their monitors. Via taskbar, this should be a quick, easy thing.
Proposed spot, context menu inside PowerToys's systray icon
Edit from @Aaron-Junker: Also add shortcuts to do these actions.
**Features**
- Adjust Resolution per monitor (#27)
- set primary monitor (#727)
- have it be a double check for "seriously, this is my primary monitor :)"
- Set brightness (#835)
- Set brightness to power state (#835)
- Adjust DPI per monitor
- Set nightmode (#48)
- Adjust based on brightness sensor (#718)
- Set HDR (#48)
- 3D mode (#48) <- is this really a thing people toggle?
- Create profiles (#718)
- Load profiles (#718)
- Save profiles (#718)
- sleep monitors | Idea-New PowerToy,Product-Display management | high | Critical |
546,142,082 | pytorch | `index_add_` with multidimensional index | ## 🚀 Feature
Allow multidimensional index in `index_add_`
## Motivation
`index_add_` is similar to `scatter_add_` with the index applied on RHS rather than LHS. Unfortunately, currently `scatter_add_` allows multidimensional index while `index_add_` allows 1D index only.
Allowing multidimensional index can make this function much more useful. For example, if a problem involves some kind of (batch) sorting, one might need to permute another tensor the same way as the sorted tensor. In this case, the 1D `index_add_` becomes useless and explicit advanced indexing will waste time doing copy.
I see a similar issue #30574. This is kind of different compared to that because that is more-or-less an api that converts advanced indexing to proper language, while this is a request of a fused-operation.
## Pitch
Make `index_add_` exactly the same as `scatter_add_`, except that the index is applied on the RHS, like:
```python
self[i][j][k] += src[index[i][j][k]][j][k] # if dim == 0
self[i][j][k] += src[i][index[i][j][k]][k] # if dim == 1
self[i][j][k] += src[i][j][index[i][j][k]] # if dim == 2
```
## Alternatives
The alternative is directly do advanced indexing:
```python
self += src[index, arange(src.size(1))[:,None], arange(src.size(2))]
self += src[arange(len(src))[:,None,None], index, arange(src.size(2))]
self += src[arange(len(src))[:,None,None], arange(src.size(1))[:,None], index]
```
But this creates permuted copy of `src` and is less efficient.
CC @ngimel | triaged,module: advanced indexing,function request | low | Major |
546,160,271 | TypeScript | Returned tuples don't match compatible types when using function union types | The first error is especially odd.
**TypeScript Version:** 3.7.x-dev.20200104
<!-- Search terms you tried before logging this (so others can find this issue more easily) -->
**Search Terms:**
**Code**
```ts
type C1 = (a: 1) => [0];
type C2 = (a: 2) => [number];
type C3 = (a: 3) => undefined;
declare function f(a: C1 | C3): void;
f(() => [0]);
declare function g(a: C2 | C3): void;
g(() => [0]);
```
**Expected behavior:**
pass
**Actual behavior:**
Type 'number' is not assignable to type '0'.(2322)
Property '0' is missing in type 'number[]' but required in type '[number]'.(2741)
**Playground Link:** http://www.typescriptlang.org/play/index.html?ts=3.8.0-dev.20200104&ssl=1&ssc=1&pln=8&pc=14#code/C4TwDgpgBAwgjFAvFAFAQwFxTgSiQPigG0AGAXQG4AoUSWAJiVUynr0UKIDsBXAWwBGEAE6Ua4aDADMTdFintCPLgBMIAMwCWXCCupU1AYwA2aYdHXLDwTQHsuUdXNgIAPrAVYAbrc16qTiiKxOQ41Eam5o5WNvZQAObOMIzu0jjevv6JQQQhZGFAA
**Related Issues:** <!-- Did you find other bugs that looked similar? -->
| Bug | low | Critical |
546,224,521 | youtube-dl | MTV.it | <!--
######################################################################
WARNING!
IGNORING THE FOLLOWING TEMPLATE WILL RESULT IN ISSUE CLOSED AS INCOMPLETE
######################################################################
-->
## Checklist
<!--
Carefully read and work through this check list in order to prevent the most common mistakes and misuse of youtube-dl:
- First of, make sure you are using the latest version of youtube-dl. Run `youtube-dl --version` and ensure your version is 2020.01.01. If it's not, see https://yt-dl.org/update on how to update. Issues with outdated version will be REJECTED.
- Make sure that all provided video/audio/playlist URLs (if any) are alive and playable in a browser.
- Make sure that site you are requesting is not dedicated to copyright infringement, see https://yt-dl.org/copyright-infringement. youtube-dl does not support such sites. In order for site support request to be accepted all provided example URLs should not violate any copyrights.
- Search the bugtracker for similar site support requests: http://yt-dl.org/search-issues. DO NOT post duplicates.
- Finally, put x into all relevant boxes (like this [x])
-->
- [X] I'm reporting a new site support request
- [X] I've verified that I'm running youtube-dl version **2020.01.01**
- [X] I've checked that all provided URLs are alive and playable in a browser
- [X] I've checked that none of provided URLs violate any copyrights
- [X] I've searched the bugtracker for similar site support requests including closed ones
## Example URLs
<!--
Provide all kinds of example URLs support for which should be included. Replace following example URLs by yours.
-->
- Single video: http://www.mtv.it/video/vkegna/Mario-Una-Serie-Di-Maccio-Capatonda-3x18-Best-of-2
- Single video: http://www.mtv.it/video/p1nyn4/wochit-selena-gomez-look-da-copiare
- Single video: http://www.mtv.it/video/39z50q/Mario-Una-Serie-Di-Maccio-Capatonda-3x17-Best-of-1
- Single video: http://www.mtv.it/episodi/8lhwng/mario-una-serie-di-maccio-capatonda-penultima-puntata-episodio-15-Sn-a-E15
- Single video: http://www.mtv.it/video/topics/Musica/nn37kq/iom2019-zorzi-ava
## Description
<!--
Currently youtube-dl for those links downloads an HTML page.
Seems that the stream is scrambled using a key that is obtained initially. I'm opening this as an issue here because probably you have more experience than me on what to do with the key and the stream.
-->
MTV.it is not working. It downloads just some HTML. | site-support-request | low | Critical |
546,253,814 | terminal | Incorrect Font Fallback Behavior (we fall back to Segoe UI sometimes!) | _notes from @DHowett_ - we're also falling back to `Segoe UI Symbol` when box-drawing glyphs don't exist. We should have a smart internal font fallback table that we consult _even when we allow customization in #2664._
<!--
This bug tracker is monitored by Windows Terminal development team and other technical folks.
**Important: When reporting BSODs or security issues, DO NOT attach memory dumps, logs, or traces to Github issues**.
Instead, send dumps/traces to [email protected], referencing this GitHub issue.
If this is an application crash, please also provide a Feedback Hub submission link so we can find your diagnostic data on the backend. Use the category "Apps > Windows Terminal (Preview)" and choose "Share My Feedback" after submission to get the link.
Please use this form and describe your issue, concisely but precisely, with as much detail as possible.
-->
# Environment
```none
Microsoft Windows [Version 10.0.18362.535]
Windows Terminal version (if applicable): 0.7.3451.0
```
# Steps to reproduce
Print anything in Hebrew when using a font without Hebrew glyphs (e.g. Cascadia Code or Consolas).
# Expected behavior
So, leaving aside RTL being tough, there's a very specific error here. Namely, font fallback in the code seems to be:
Consolas --> Lucida Console --> Courier New.
This should fail the first two, but the Hebrew charset **is** covered in Courier New, so that should be the font used here. If you set the font to Courier New, you can see that the font rendering itself is not messed up. Still RTL is tough, but -- the ransom note aspect of it is gone.
# Actual behavior
With the benefit of super zooming in, it seems to me that the font being used for Hebrew is *Segoe UI*. I assume that it's some sort of system default, MSFT and whatnot, but that's really not good. Courier New may be jarring, as you'd expect from fallback, but given that it's at least monospaced I assume that the ransom note aspect of the text rendering here wouldn't happen. Segoe UI is not monospaced, and I think that's why you get the ransom note like variation in text sizes.
An important clarification / note: this is **_NOT_** a regression. This is not a new bug, it's been around forever. The only new thing is that I read the source code yesterday and realized that it wasn't supposed to happen.
<!-- What's actually happening? -->
| Area-Rendering,Issue-Bug,Product-Terminal,Priority-2 | low | Critical |
546,273,923 | go | x/crypto/blake2s: add 224bit variant | <!--
Please answer these questions before submitting your issue. Thanks!
For questions please use one of our forums: https://github.com/golang/go/wiki/Questions
-->
### What version of Go are you using (`go version`)?
irrelevant
### Does this issue reproduce with the latest release?
yes
### What operating system and processor architecture are you using (`go env`)?
irrelevant
### What did you do?
Went to https://godoc.org/golang.org/x/crypto/blake2s and looked for 224bit version.
### What did you expect to see?
Something like `func New224(key []byte) (hash.Hash, error)` (kinda like in [crypto/sha256](https://golang.org/pkg/crypto/sha256/#New224)) or `func New(size int, key []byte) (hash.Hash, error)` (like in [x/crypto/blake2b](https://godoc.org/golang.org/x/crypto/blake2b#New)).
### What did you see instead?
I didn't find anything what would clearly enable using canonical blake2s-224 hash.
Truncating New256 output would be ugly and wouldn't match canonical blake2s-224 (output size is embedded into IV).
I don't think blake2s XOF would provide canonical blake2s-224 either. | NeedsInvestigation,FeatureRequest | low | Critical |
546,291,030 | flutter | Inconsistent behavior of RawKeyboardListener on different devices/Android versions | I am trying to get my flutter app to react to events generated when media keys are pressed on a little bluetooth keyboard paired with my android device.
Here's the code snippet that builds the widget I use to evaluate the scenario.
If I test it on my Google Pixel C tablet running under Android 8.1.0, kernel version 3.18.0-g8ce8444 it works as expected, When the widget is displayed I can press the media buttons on the keyboard and the events are properly generated and passed to the app, handlers are fired etc.
On the other hand, if I test exactly the same code on my Samsung S9 running under Android 9, kernel version 4.9.59-17174122 my attempts to press the media buttons on the keyboard will take no effect until I press one of the numeric keys. After that, the following lines appear in the log
After that the media keys on the keyboard work as expected.
Is there any workaround to that problem?
Finally a question regarding iOS. I am aware that RawKeyboardListerer is not implemented for iOS, but is there a way for a flutter app running on iOS device to react to media buttons like in the above scenario? | a: text input,platform-android,framework,engine,P2,team-design,triaged-design | low | Major |
546,332,298 | TypeScript | Suggestion: optional globals | <!-- 🚨 STOP 🚨 𝗦𝗧𝗢𝗣 🚨 𝑺𝑻𝑶𝑷 🚨
Half of all issues filed here are duplicates, answered in the FAQ, or not appropriate for the bug tracker.
Please help us by doing the following steps before logging an issue:
* Search: https://github.com/Microsoft/TypeScript/search?type=Issues
* Read the FAQ, especially the "Common Feature Requests" section: https://github.com/Microsoft/TypeScript/wiki/FAQ
-->
## Search Terms
<!-- List of keywords you searched for before creating this issue. Write them down here so that others can find this suggestion more easily -->
## Suggestion
Extracted from this related issue: https://github.com/microsoft/TypeScript/issues/21965
Currently TypeScript has no way to represent a global which may or may not exist at runtime.
For example, given a global called `foo`, we can only define it as:
```ts
declare global {
const foo: number;
}
```
But what if `foo` does not always exist?
We can't define it as `T | undefined` because this does not reflect the runtime behaviour. `T | undefined` means "this will be declared but its value might be `undefined`", but the global may not be declared at all, in which case any references would result in `ReferenceError`s at runtime.
```ts
declare global {
const foo: number | undefined;
}
```
```ts
// TypeScript is happy for us to reference this. There's no compile error.
// The type is `number | undefined`.
// At runtime, if the global doesn't exist, we'll get a `ReferenceError` 😒
foo;
```
If there was a way to mark a global as optional, TypeScript would require a proper guard (`typeof foo !== 'undefined'`) before it can be safely accessed. This way, TypeScript is alerting us to the possibilty of a runtime exception.
```ts
declare global {
// example syntax
const foo?: number;
}
```
```ts
// Compile error 😇
// Runtime exception 😇
foo;
if (typeof foo !== 'undefined') {
// No compile error 😇
// No runtime exception 😇
// Type is `number`
foo;
}
```
## Use Cases
Examples of "optional globals" below.
In code that is shared between a client (browser) and server (Node):
- `window` will only exist during the client runtime
```js
typeof window !== undefined ? window.alert('yay') : undefined;
```
- `global` and `require` will only exist during the server runtime
```js
typeof process !== undefined ? process.env.FOO : undefined;
```
In code which may or may not be under test using [Cypress](https://www.cypress.io/), the global `Cypress` will only exist during runtime when the code is under test.
```js
const getEnvVar = key => typeof Cypress !== 'undefined' ? Cypress.env(key) : process.env[key];
```
## Examples
See above.
## Checklist
My suggestion meets these guidelines:
* [x] This wouldn't be a breaking change in existing TypeScript/JavaScript code
* [x] This wouldn't change the runtime behavior of existing JavaScript code
* [x] This could be implemented without emitting different JS based on the types of the expressions
* [x] This isn't a runtime feature (e.g. library functionality, non-ECMAScript syntax with JavaScript output, etc.)
* [x] This feature would agree with the rest of [TypeScript's Design Goals](https://github.com/Microsoft/TypeScript/wiki/TypeScript-Design-Goals).
| Suggestion,Awaiting More Feedback | low | Critical |
546,344,574 | terminal | Polish Commandline Help MessageBox to be at least like, but at best better than, `msiexec /?` | Follow-up task from #4023
## actual
## expected
(ideally with monospace text too)
Obviously, better than this would be even more ideal. It's probably not that simple to create a XAML island just for the sake of displaying this text. However, that would be potentially even more modern UI. | Help Wanted,Area-UserInterface,Product-Terminal,Issue-Task,Priority-3,Area-Commandline,good first issue | low | Major |
546,388,290 | vscode | API to programatically expand/collapse tree view | ### Request
An API on TreeView which allows extensions to expand or collapse their contributed views within the view container. That is, expand/collapse the view itself, not the items in the view.
### Example Use Case
If I have a view container with multiple contributed views, not all views may be relevant all the time. It would be nice to be able to close the irrelevant views, to give more screen real estate to the relevant ones.
For example, maybe one view shows the definition of the currently selected word in the editor: if nothing in the editor is selected, then ideally this view could just be collapsed since it is providing no value to the user. When something is selected, it should be possible for the extension to re-expand the view.
| feature-request,api,tree-views | medium | Critical |
546,397,780 | rust | Implementation of std::net::TcpListener::shutdown | Is there any reason why it does not exist yet besides putting in the effort of writing it? | T-libs-api,C-feature-request | low | Major |
546,406,254 | svelte | <input> elements in an {#if} that depends on the input value do not fire on:input before they disappear | **Describe the bug**
When you create an input that can be removed because of an if clause around it that depends on the input value, the last `on:input` event of the <input> is not fired because it seems to disappear too quickly.
**Logs**
**To Reproduce**
Minimal example: Remove the text in the input field character by character. For the last input event (the last BackSpace keystroke), no event is fired and `onInput` is not called.
https://svelte.dev/repl/8d8bde660dea46f6b5443fe701cabd2b?version=3.16.7
**Expected behavior**
Before the input field disappears, one last input event is fired and `value: ` is output into the console.
**Information about your Svelte project:**
- Browser: Google Chrome version 79.0.3945.88
- Operating System: Windows 10
- Svelte version: 3.16.7 (newest release)
- Webpack or Rollup: not needed to reproduce the bug (REPL pad)
**Severity**
Not so high: It's just a bit confusing, as I would expect the last input event to fire. This can easily be fixed manually by using `display: none` CSS instead of a Svelte `{#if}` clause.
Manual fix with `display: none` demo (also demonstrating the expected behavior): https://svelte.dev/repl/78fd0c98c9cf4f3fab2c7e6f43cf7fe3?version=3.16.7
| feature request,temp-stale,documentation | low | Critical |
546,406,535 | go | all: stop using non-libc system calls on OpenBSD | Upcoming changes to the OpenBSD kernel will prevent system calls from being made unless they are coming from libc.so (with some exceptions, for example, a static binary). There are also likely to be changes to the APIs used for system calls. As such, the Go runtime (and other packages) need to stop using direct syscalls, rather calling into libc functions instead (as has always been done for Solaris and now also for macOS). This will avoid these issues and allow Go to continue to function correctly on OpenBSD.
A version of Go with openbsd/amd64 being modified to perform all calls via libc is available at:
https://github.com/4a6f656c/go/tree/openbsd-syscall
Further work is still required to convert openbsd/386, openbsd/arm and openbsd/arm64. | help wanted,OS-OpenBSD,NeedsFix | high | Critical |
546,409,387 | go | net: error message for ListenIP is confusing | # net.ListenIP
The documentation of `net.ListenIP` says:
```
The network must be an IP network name; see func Dial for details.
```
However using one of the documented IP network names (like "ip" or "ip4") causes the function to return an error on Linux; e.g. `unknown network ip`.
According to #7439:
```
The first argument of ListenIP should be "network"+":"+"protocol on top of the IPv4 or
IPv6".
````
I confirmed that this works on Linux, but the function is clearly not documented correctly. From #7439 I assume that the format of `network` is platform dependant.
# net.DialIP
The documentation of the network parameter of `net.DialIP` has the same problem as in `net.ListenIP`. | Documentation,NeedsInvestigation | low | Critical |
546,417,762 | pytorch | I'm not able to build pytorch with tensorrt (current master) | ## Issue description
I'm not able to build pytorch with tensorrt (current master)
```
In file included from ../caffe2/contrib/tensorrt/tensorrt_tranformer.cc:16:0:
../caffe2/opt/backend_cutting.h:5:10: fatal error: nomnigraph/Representations/NeuralNet.h: No such file or directory
#include "nomnigraph/Representations/NeuralNet.h"
```
## System Info
```
Collecting environment information...
PyTorch version: N/A
Is debug build: N/A
CUDA used to build PyTorch: N/A
OS: Ubuntu 18.04.1 LTS
GCC version: (Ubuntu 7.4.0-1ubuntu1~18.04.1) 7.4.0
CMake version: version 3.14.0-rc2
Python version: 3.7
Is CUDA available: N/A
CUDA runtime version: 10.2.89
GPU models and configuration:
GPU 0: GeForce GTX 1050 Ti
GPU 1: TITAN RTX
Nvidia driver version: 440.44
cuDNN version: /usr/lib/x86_64-linux-gnu/libcudnn.so.7.4.2
Versions of relevant libraries:
[pip] numpy==1.18.0
[pip] numpydoc==0.9.1
[conda] blas 1.0 mkl
[conda] mkl 2019.4 243
[conda] mkl-service 2.3.0 py37he904b0f_0
[conda] mkl_fft 1.0.14 py37ha843d7b_0
[conda] mkl_random 1.1.0 py37hd6b4f25_0
```
- PyTorch or Caffe2: pytorch
- How you installed PyTorch (conda, pip, source): source
- Build command you used (if compiling from source): make
- OS: Ubuntu
- PyTorch version: git hash 20c5dd5
- Python version: 3.7
- CUDA/cuDNN version: 10.2 / 7
- Versions of any other relevant libraries: tensorrt 7
| module: build,triaged | low | Critical |
546,426,016 | flutter | Consider copying MultiChildRenderObjectWidget.children | One of the problems with the Flutter model is that developers can accidentally or intentionally mutate the Widget tree via mutating MultiChildRenderObjectWidget.children. See https://github.com/dart-lang/sdk/issues/27755.
I think there might be a performance reason to copy the children. It depends on how often the framework walks the `children`.
`List` is an interface that has many implementations.
The VM has `_List` for fixed-length lists and `_GrowableList` for 'regular' lists, and another implementation (I don't recall the name) for unmodifiable and const `List`s.
In a small Flutter app, all the children are likely from `List` literals or to calls to `.toList()`, so the `children` field is likely to have just two or three implementation classes. The JITing VM and AOT try to discover this, and the three implementation classes are carefully crafted to be as alike as possible. This makes it possible to compile `children[i]` to some indexing code.
It is possible that in a large Flutter app that there are other implementations of `List` that are passed in as the children, meaning it is no longer possible to compile `children[i]` to indexing code since some of the implementations have custom indexers and `.length` getters that must be called.
My hypothesis is that as the inner loop in the framework becomes more polymorphic, the performance of this code shared by all multi-child widgets degrades.
One could test this hypothesis by taking a benchmark that is where walking the children is 'hot'. I expect the benchmark would walk the 'the same' a large Widget tree over and over, with no actual appearance change. Alter one of the constructors to pass in an `UnmodifiableListView([....])` instead of `[....]`. This will cause the walking of `children[i]` to become less efficient. Does this show up in the benchmark?
If you decide to protect the performance by copying (which ensures there is only one kind of List), I would suggest copying with `List<Widget>.unmodifiable(input)`.
A small fixed-length or unmodifiable list can be is half the size of a growable list or less. If the growable input is no longer referenced, it can be GC-ed. The copy will do more allocation, but have a smaller size post-GC. This may be a reasonable tradeoff.
Making the children unmodifiable will prevent users from modifying the trees they create which will help them avoid a class of bugs.
It will break the problematic pattern of `widget.children.add(w)`.
Even if the above hypothesis does not pan out, copying the list to an unmodifiable list in debug mode would help developers avoid bugs.
| framework,c: performance,c: proposal,perf: speed,P2,team-framework,triaged-framework | low | Critical |
546,443,548 | vscode | Ability to drag and drop breakpoints | Issue Type: <b>Feature Request</b>
The ability to move breakpoints, especially those that are conditional, would be considerably easier than having to delete a breakpoint and then recreate it with the same condition information on another line. Drag and drop functionality would work pretty well here I imagine.
VS Code version: Code 1.41.1 (26076a4de974ead31f97692a0d32f90d735645c0, 2019-12-18T14:58:56.166Z)
OS version: Windows_NT x64 10.0.18362
<!-- generated by issue reporter --> | feature-request,debug | medium | Critical |
546,446,233 | rust | ICE in unsized fn params: tried to statically allocate unsized place | The following ICE's in nightly:
```Rust
#![feature(unsized_fn_params)]
fn main() {
let f: fn([u8]) = |_| {};
let slice: Box<[u8]> = Box::new([1; 8]);
f(*slice);
}
```
<details>
<summary>Backtrace:</summary>
```
Compiling playground v0.0.1 (/playground)
thread 'rustc' panicked at 'tried to statically allocate unsized place', src/librustc_codegen_ssa/mir/place.rs:55:9
stack backtrace:
0: backtrace::backtrace::libunwind::trace
at /cargo/registry/src/github.com-1ecc6299db9ec823/backtrace-0.3.40/src/backtrace/libunwind.rs:88
1: backtrace::backtrace::trace_unsynchronized
at /cargo/registry/src/github.com-1ecc6299db9ec823/backtrace-0.3.40/src/backtrace/mod.rs:66
2: std::sys_common::backtrace::_print_fmt
at src/libstd/sys_common/backtrace.rs:77
3: <std::sys_common::backtrace::_print::DisplayBacktrace as core::fmt::Display>::fmt
at src/libstd/sys_common/backtrace.rs:59
4: core::fmt::write
at src/libcore/fmt/mod.rs:1057
5: std::io::Write::write_fmt
at src/libstd/io/mod.rs:1426
6: std::sys_common::backtrace::_print
at src/libstd/sys_common/backtrace.rs:62
7: std::sys_common::backtrace::print
at src/libstd/sys_common/backtrace.rs:49
8: std::panicking::default_hook::{{closure}}
at src/libstd/panicking.rs:195
9: std::panicking::default_hook
at src/libstd/panicking.rs:215
10: rustc_driver::report_ice
11: std::panicking::rust_panic_with_hook
at src/libstd/panicking.rs:476
12: std::panicking::begin_panic
13: rustc_codegen_ssa::mir::place::PlaceRef<V>::alloca
14: <core::iter::adapters::Map<I,F> as core::iter::traits::iterator::Iterator>::fold
15: <alloc::vec::Vec<T> as alloc::vec::SpecExtend<T,I>>::from_iter
16: rustc_codegen_ssa::mir::codegen_mir
17: <rustc::mir::mono::MonoItem as rustc_codegen_ssa::mono_item::MonoItemExt>::define
18: rustc_codegen_llvm::base::compile_codegen_unit::module_codegen
19: rustc::dep_graph::graph::DepGraph::with_task
20: rustc_codegen_llvm::base::compile_codegen_unit
21: rustc_codegen_ssa::base::codegen_crate
22: <rustc_codegen_llvm::LlvmCodegenBackend as rustc_codegen_utils::codegen_backend::CodegenBackend>::codegen_crate
23: rustc_interface::passes::start_codegen::{{closure}}
24: rustc::util::common::time
25: rustc_interface::passes::start_codegen
26: rustc::ty::context::tls::enter_global
27: rustc_interface::queries::Queries::ongoing_codegen
28: rustc_interface::interface::run_compiler_in_existing_thread_pool
29: scoped_tls::ScopedKey<T>::set
30: syntax::with_globals
note: Some details are omitted, run with `RUST_BACKTRACE=full` for a verbose backtrace.
error: internal compiler error: unexpected panic
note: the compiler unexpectedly panicked. this is a bug.
note: we would appreciate a bug report: https://github.com/rust-lang/rust/blob/master/CONTRIBUTING.md#bug-reports
note: rustc 1.42.0-nightly (da3629b05 2019-12-29) running on x86_64-unknown-linux-gnu
note: compiler flags: -C codegen-units=1 -C debuginfo=2 --crate-type bin
note: some of the compiler flags provided by cargo are hidden
query stack during panic:
end of query stack
error: could not compile `playground`.
```
</details> | A-codegen,T-compiler,C-bug,requires-nightly,F-unsized_fn_params | low | Critical |
546,465,702 | flutter | Flutter Inspect Widget tool does not work well with overlays | When using overlays, like Dialogs, Drawers, etc, the Select Widget mode of the inspector selects the underlying widget and not the actual widget clicked. This has always been the case, at least since Public Beta 3.
Here's a gif of it in action. As you can see, it highlighted the underlying `BottomNavigationBarItem` instead of the `Text` widget or the Drawer
> Flutter 1.13.7-pre.24 • channel master • https://github.com/flutter/flutter.git
> Framework • revision 0aed0b61a1 (4 days ago) • 2020-01-03 17:53:54 -0800
> Engine • revision eb139936eb
> Tools • Dart 2.8.0 (build 2.8.0-dev.0.0 2f57602411)
| framework,f: inspector,has reproducible steps,P2,found in release: 3.3,found in release: 3.6,team-framework,triaged-framework | low | Major |
546,466,343 | TypeScript | Allow .d.ts files to represent anonymous class expressions with `private` members | Re-opening of #30355
That other issue is labeled "working as intended" so instead of this feature being a bug report, it is a feature request. | Suggestion,Awaiting More Feedback | medium | Critical |
546,478,055 | flutter | Add orElse or a fallback to Navigator.popUntil() | <!-- Thank you for using Flutter!
If you are looking for support, please check out our documentation
or consider asking a question on Stack Overflow:
* https://flutter.dev/
* https://api.flutter.dev/
* https://stackoverflow.com/questions/tagged/flutter?sort=frequent
If you have found a bug or if our documentation doesn't have an answer
to what you're looking for, then fill our the template below. Please read
our guide to filing a bug first: https://flutter.dev/docs/resources/bug-reports
-->
## Use case
I would like to add some more stability to my app when navigating back to the login screen after logging out. Instead of ensuring that a screen is on the stack, which I can't do programmatically (since we don't have external access to the stack yet, (#45938 maybe?), I want to do popUntil(ModalRoute.named('/') however, in case this screen doesn't exist in the stack, maybe the user was deep-linked in or something, I want to have a fallback and provide it anyway.
<!--
Please tell us the problem you are running into that led to you wanting
a new feature.
Is your feature request related to a problem? Please give a clear and
concise description of what the problem is.
Describe alternative solutions you've considered. Is there a package
on pub.dev/flutter that already solves this?
-->
## Proposal
Add the ability to check to see if a certain route is in the stack, or allow us to implement a fallback screen in the case of this call returning unsuccessfully.
<!--
Briefly but precisely describe what you would like Flutter to be able to do.
Consider attaching images showing what you are imagining.
Does this have to be provided by Flutter directly, or can it be provided
by a package on pub.dev/flutter? If so, maybe consider implementing and
publishing such a package rather than filing a bug.
-->
| c: new feature,framework,f: routes,c: proposal,P3,team-framework,triaged-framework | low | Critical |
546,487,415 | go | cmd/go: reject vN.*.*+incompatible as a version if vN/go.mod exists | (From a [golang-nuts thread](https://groups.google.com/d/topic/golang-nuts/eRt5gD5pIhQ/discussion).)
### What version of Go are you using (`go version`)?
<pre>
~$ go version
go version devel +7d98da8d31 Tue Jan 7 19:53:19 2020 +0000 linux/amd64
</pre>
### What did you do?
Run `go get` in module mode, requesting an incompatible major version of a module that _has_ a `go.mod` file corresponding to that major version _using the “major subdirectory” layout_, but lacks a `go.mod` file altogether at the repo root.
```
~$ go clean -modcache
~$ GOPROXY=direct GO111MODULE=on go get -d github.com/nicksnyder/[email protected]
go: downloading github.com/nicksnyder/go-i18n v2.0.3+incompatible
go: github.com/nicksnyder/go-i18n v2.0.3 => v2.0.3+incompatible
```
### What did you expect to see?
An error message indicating that `v2.0.3+incompatible` is not valid for `github.com/nicksnyder/go-i18n` because the repository contains a `v2/go.mod`.
(See #32695 and #34165.)
### What did you see instead?
```
~$ GOPROXY=direct GO111MODULE=on go get -d github.com/nicksnyder/[email protected]
go: downloading github.com/nicksnyder/go-i18n v2.0.3+incompatible
go: github.com/nicksnyder/go-i18n v2.0.3 => v2.0.3+incompatible
```
CC @matloob @jayconrod @thepudds @heschik @hyangah @katiehockman | NeedsDecision,modules | low | Critical |
546,536,836 | go | cmd/compile: manually CSE |-combined rules? | We know that rewrite rules containing `|` have a lot of shared subexpressions. Instead of expanding and generating one rewrite rule per expanded rule, we could instead loop over all op tuples, much as CL 213703 proposes we do for commutative ops. Note that this only works for |-expansions that aren't for the top level op, because the top level op is matched using a switch statement. Perhaps surprisingly, about 15% of |-expanded rules (both before and after expansion) use | only in non-top-level rules.
| NeedsInvestigation,compiler/runtime | low | Minor |
546,574,646 | rust | Incompatible type error should explain when types come from default type parameters | ```rust
trait Trait<T = ()> {
fn foo() -> T;
}
struct S;
impl Trait for S {
fn foo() -> u32 { 0 }
}
```
produces:
```
error[E0053]: method `foo` has an incompatible type for trait
--> src/lib.rs:28:17
|
22 | fn foo() -> T;
| - type in trait
...
28 | fn foo() -> u32 { 0 }
| ^^^ expected `()`, found `u32`
|
= note: expected fn pointer `fn()`
found fn pointer `fn() -> u32`
```
The diagnostic doesn't point out why `T` is `()`, which comes from the default type parameter. It would be helpful to explain this, as it may not be immediately apparent. | C-enhancement,A-diagnostics,T-compiler,D-confusing | low | Critical |
546,585,039 | TypeScript | Either understand const arrow assertions or give better errors on them | As mentioned by @threehams in https://github.com/microsoft/TypeScript/issues/35838#issuecomment-569832017, people are pretty confused about the rules for assertion functions, and I don't blame them. For example
```ts
const assert = (blah: unknown): asserts blah => { throw "hai"; }
let x: string | undefined;;
assert(x);
```
In our nightly releases, we give an elaboration on the declaration of `assert` like
```
'assert' needs an explicit type annotation.
```
For all intents and purposes, it looks like `assert` *does* have an annotation! The problem is that `assert` itself needs a `: (x: unknown) => asserts x`, because it's not automatically inferred from the arrow function.
We have two options:
* Remove this restriction when the declaration of an assertion function is trivially detectable.
* Make assertion function errors more clear for arrow functions:
```
'assert' needs an explicit type annotation. While the function it has been assigned to is fully annotated, the variable declaration for 'assert' does not.
``` | Suggestion,In Discussion,Domain: Error Messages,Experience Enhancement,Rescheduled | medium | Critical |
546,616,354 | pytorch | RPC and dist_autograd should respect no_grad mode | It might be helpful to allow using RPCs to do some non-autograd actions within an autograd context. For example, it will be useful to support the following use case:
```python
with dist_autograd.context():
# this RPC should add send/recv functions
rpc_sync(dst, some_func, args=(some_tensor))
with torch.no_grad():
# this RPC should NOT add send/recv functions or carry context id
rpc_sync(dst, some_other_func, args=(some_tensor))
# this RPC should add send/recv functions
rpc_sync(dst, some_func, args=(some_tensor))
```
cc @pietern @mrshenli @pritamdamania87 @zhaojuanmao @satgera @gqchen @aazzolini @rohan-varma @xush6528 @jjlilley | feature,triaged,module: rpc | low | Minor |
546,623,569 | opencv | Are there plans to support .NET Core? | RT | feature,priority: low | low | Minor |
546,624,024 | flutter | Difficulties when implementing custom text layout for vertical Mongolian script | Since issue #35994 was closed without being solved, this issue an attempt to follow the [advice](https://github.com/flutter/flutter/issues/35994#issuecomment-571277992) from @Hixie to open a specific use case example and describe the bugs that prevent it from being implemented.
## What I am trying to do
Take Mongolian text that would be rendered as follows in a normal Text widget:
And render it vertically with left to right line wrapping. Note that the circled characters in the above image should maintain the same orientation.
<img width="212" alt="mongolian_layout_2" src="https://user-images.githubusercontent.com/10300220/61083462-72cd1700-a3f1-11e9-91e3-9a90a734e373.png">
## Why the current tools don't work
- Doing a simple rotation of a Text widget doesn't work because the lines should wrap from left to right.
- `LineMetrics`, which I reviewed [here](https://medium.com/swlh/flutter-line-metrics-fd98ab180a64), doesn't provide an easy way to access the text of a line. But even if it did, creating a new Paragraph for each line and drawing it in the correct location and orientation isn't enough because emoji and CJK characters within the line need to be rotated differently than the Latin and Mongolian text in that line.
- To know how much text will fit in a line, each word and emoji/CJK character has to be measured. There is no current tool to know where the word and character breaks occur within the string. A naive approach is to break at the spaces, but there are many other places that breaks could and should appear.
- It is not possible to measure or draw a substring within a Paragraph. The only way I know how to do it is to create a new Paragraph object for every single text run (ie, word/emoji/CJK character). This can easily lead to hundreds or even thousands of Paragraphs. I haven't done testing on performance myself yet, but [performance issues have been reported with this](https://github.com/flutter/flutter/issues/35994#issuecomment-513977291).
- This is not a bug per se, but calling a single word a `Paragraph` makes it look like the class could have a better name. It also makes the use of Paragraphs to measure and draw text feel like a hack rather than a good programming practice.
## Final thoughts
In this issue I am not suggesting specific ways I think this should be fixed in order to avoid [proposing a complex solution when a simpler one would do](https://github.com/flutter/flutter/issues/35994#issuecomment-571431454).
I'm also a little concerned that having such a specific use case will make it difficult for other people to get behind. I think when the issue is generalized a little more, there are a lot of other text rendering use cases that could be solved. But anyway, I'll start here.
Even though I've been dissatisfied with Flutter's text rendering capabilities, I do appreciate all the work the Flutter team has put into the platform as a whole and also how they are willing to listen to our requests. | c: new feature,framework,engine,a: internationalization,a: typography,c: proposal,P2,team-engine,triaged-engine | low | Critical |
546,625,306 | go | encoding/csv: writer.UseCRLF will change \n to \r\n in data field | <!--
Please answer these questions before submitting your issue. Thanks!
For questions please use one of our forums: https://github.com/golang/go/wiki/Questions
-->
### What version of Go are you using (`go version`)?
<pre>
$ go version
go version go1.13.5 linux/amd64
</pre>
### Does this issue reproduce with the latest release?
yes
### What operating system and processor architecture are you using (`go env`)?
<details><summary><code>go env</code> Output</summary><br><pre>
$ go env
GO111MODULE=""
GOARCH="amd64"
GOBIN=""
GOCACHE="/home/secret/.cache/go-build"
GOENV="/home/secret/.config/go/env"
GOEXE=""
GOFLAGS=""
GOHOSTARCH="amd64"
GOHOSTOS="linux"
GONOPROXY=""
GONOSUMDB=""
GOOS="linux"
GOPATH="/home/secret/Dropbox/go"
GOPRIVATE=""
GOPROXY="https://proxy.golang.org,direct"
GOROOT="/usr/lib/go-1.13"
GOSUMDB="sum.golang.org"
GOTMPDIR=""
GOTOOLDIR="/usr/lib/go-1.13/pkg/tool/linux_amd64"
GCCGO="gccgo"
AR="ar"
CC="gcc"
CXX="g++"
CGO_ENABLED="1"
GOMOD=""
CGO_CFLAGS="-g -O2"
CGO_CPPFLAGS=""
CGO_CXXFLAGS="-g -O2"
CGO_FFLAGS="-g -O2"
CGO_LDFLAGS="-g -O2"
PKG_CONFIG="pkg-config"
GOGCCFLAGS="-fPIC -m64 -pthread -fmessage-length=0 -fdebug-prefix-map=/tmp/go-build843557951=/tmp/go-build -gno-record-gcc-switches"
</pre></details>
### What did you do?
<!--
If possible, provide a recipe for reproducing the error.
A complete runnable program is good.
A link on play.golang.org is best.
-->
trying to write
```
"col1","col2"
"asd\njk", "2g9"
```
into csv file
but the newline in `asd\njk` has been change to `asd\r\njk`
[playground](https://play.golang.org/p/Of3BwsmGH8i)
### What did you expect to see?
`\n` in data field would not be changed by `writer.UseCRLF`
`"col1,col2\r\n\"asd\njk\",2g9\r\n"`
### What did you see instead?
`"col1,col2\r\n\"asd\r\njk\",2g9\r\n"` | NeedsInvestigation | low | Critical |
546,649,524 | vue-element-admin | 设置了 mock ,但是请求的 url 上没有包含 mock 这个字符串 | ## Question（提问）
<!--
提问之前，请确定你已经过自己的努力，尝试解决过这个问题。
若是代码相关问题，请不要只截图，请提供在线 demo，以便节约彼此的时间。
Before asking a question, please make sure that you have tried your best to solve this problem.
If it's a code-related issue, please don't just take screenshots. Please provide an online demo to save each other's time.
-->
#### Steps to reproduce（问题复现步骤）
<!--
1. [xxx]
2. [xxx]
3. [xxxx]
-->
#### Screenshot or Gif（截图或动态图）
#### Link to minimal reproduction（最小可在线还原demo）
<!--
Please only use Codepen, JSFiddle, CodeSandbox or a github repo
-->
#### Other relevant information（格外信息）
- Your OS:
- Node.js version:
- vue-element-admin version:
| enhancement :star: | low | Minor |
546,656,731 | flutter | Use a system default font on flutter | ## Use case
Flutter apps need to use a system font for displaying a plain text. I guess I can use FontLoader to use a default system font on Android with a plugin but it would be really great to have this feature supported by engine.
Samsung, LG or other vendors using Android has a feature that users can change a system font in settings but this default font setting is ignored in flutter apps because it uses a built-in font provided by flutter engine. Users using a different font as their default font feel awkward as their setting isn't applied on their Android devices.
| c: new feature,framework,f: material design,a: typography,customer: crowd,c: proposal,dependency: android,P2,team-design,triaged-design | low | Critical |
546,664,552 | excalidraw | Hint: Press shift to move elements faster with keyboard | When we move an element with keyboard, it is slow. Providing some way to speed it will be useful for power users. | enhancement | low | Major |
546,677,361 | angular | createAngularJSTestingModule() breaks tests when used multiple times across unrelated spec files | # 🐞 bug report
### Affected versions
This bug is present since the very beginning (March, 22th, 2019) when `createAngularJSTestingModule()` was first added to `@angular/upgrade/static/testing` (there have only been two commits in that file, `angular\packages\upgrade\static\testing\src\create_angularjs_testing_module.ts`).
This problem only occurs (I suppose) with Jasmine. When using Jest, I think the test runs are already isolated so the problem might not show up there.
### Description
Calling `createAngularJSTestingModule()` immediately calls (under the hood) `angular.module(...)`, thus it "globally" adds the required angularJSTestingModule to the "global storage" of all angularJS modules. This works fine so long as you always require the same `angularModules` (like in `createAngularJSTestingModule(angularModules)`). However, once you have multiple test suites (i.e. different and totally unrelated spec files), you'll start to require different angularModules from time to time.
Now, due to `createAngularJSTestingModule()` immediately calling `angular.module()`, all previous invocations get overriden. "In the end", only the last call "wins" with its required angularModules. All previous calls are in vane, making their respective test suites fail, because they don't have (or load) the required `angularModules`.
To understand why this happens, you need to recall that Jasmine will first execute ALL describe blocks for ALL test files that are part of the test run (I guess this is needed so that Jasmine can provide the "focused specs" feature). Thus, when you stick to the notation provided in the official docs, you're lost:
A workaround for the problem is to not call `angular.mock.module()` directly, but provide a function to `beforeEach` instead (which will only be called later, just before the test is run):
```
beforeEach(() => angular.mock.module(createAngularJSTestingModule(angularModules))); // notice the "() =>" to the left
```
The simplest solution would be to update the docs accordingly. However, when you're using `angular.mock.module.sharedInjector()` you'll most likely have a `beforeAll` instead of an `beforeEach`. While the workaround looks the same (just replace the beforeEach with beforeAll), you can no longer use a `beforeAll` in combination with `inject()`. Those calls need to be replaced instead with a `beforeEach`. This is due to `beforeAll(inject(...))` creating the injector too early which makes TestBed feel uncomfortable (erroring).
Maybe there's yet another, better, solution?
When looking at the source code:
Maybe it's feasible to use an internal counter to create different modules each time?
Or maybe it's possible to not call `angular.module()` at all, and provide a simple "module function" instead, that will be unique to the test suite?
@petebacondarwin | type: bug/fix,workaround1: obvious,area: upgrade,state: confirmed,P5 | low | Critical |
546,713,297 | rust | Spurious and unrelated error when using a macro variable (expr) | After some more consideration I'm starting to think this is an issue with rust's macro system rather than any of the used macros; see https://github.com/rust-lang/futures-rs/issues/2026.
Note the `($should_succeed:expr)`. If `$should_succeed` is used at all in the block starting at line 85, I get an error that `id` is not found, which seems totally random and unrelated.
However I just found that if I instead do `($should_succeed:tt)` I get no such error. | A-macros,T-compiler,C-bug | low | Critical |
546,736,765 | pytorch | Support sparse inputs for torch.block_diag | ## 🚀 Feature
The blkdiag method is defined clearly in https://github.com/pytorch/pytorch/issues/31932
https://github.com/pytorch/pytorch/issues/31932 suggests blkdiag should create a dense Tensor, which may also be helpful in some case.
However, considering graph neural networks, we always want a sparse block tensor rather than a dense one, since a dense block tensor will be even slower than multiply submatrix one by one and will easily cause OOM.
A use case can be found in https://stackoverflow.com/a/59641321/7699035
It's consistent with the most popular pytorch_geometric module, where node features x1, x2, x3, ..., xn of different graphs are concatenated to a large tensor and a batch index is given. I've also asked the author of pytorch_geometric on this problem here https://github.com/rusty1s/pytorch_scatter/issues/95 .
## Pitch
This issue is for something like torch.spase.blkdiag rather than torch.blkdiag.
## Alternatives
The operation in https://stackoverflow.com/a/59641321/7699035 is clearly parallelizable, I want an efficient solution in pytorch, however a torch.sparse.blkdiag method seems the best solution.
cc @vincentqb @aocsa @nikitaved @pearu @mruberry | module: sparse,triaged,module: tensor creation,function request | low | Major |
546,740,715 | node | Inspector does not initialize when using CreateEnvironment | Inspector does not initialize when using CreateEnvironment.
So occur crash in the code below. (inspector_agent()->client_ is null)
> https://github.com/nodejs/node/blob/1a552f614b4ec2b9faf0bf932ad4e6ef4fdcaf01/src/node_worker.cc#L79-L80
> https://github.com/nodejs/node/blob/3e79c004fdb93d01618fd90f0df934ac12c62353/src/inspector_agent.cc#L1010
Initialization takes place only here.
> https://github.com/nodejs/node/blob/1a552f614b4ec2b9faf0bf932ad4e6ef4fdcaf01/src/node_main_instance.cc#L225-L231
```c++
Environment* CreateEnvironment(IsolateData* isolate_data,
Local<Context> context,
int argc,
const char* const* argv,
int exec_argc,
const char* const* exec_argv) {
Isolate* isolate = context->GetIsolate();
HandleScope handle_scope(isolate);
Context::Scope context_scope(context);
// TODO(addaleax): This is a much better place for parsing per-Environment
// options than the global parse call.
std::vector<std::string> args(argv, argv + argc);
std::vector<std::string> exec_args(exec_argv, exec_argv + exec_argc);
// TODO(addaleax): Provide more sensible flags, in an embedder-accessible way.
Environment* env = new Environment(
isolate_data,
context,
args,
exec_args,
static_cast<Environment::Flags>(Environment::kIsMainThread |
Environment::kOwnsProcessState |
Environment::kOwnsInspector));
env->InitializeLibuv(per_process::v8_is_profiling);
// ========== START SUGGEST ==========
env->InitializeDiagnostics();
// TODO(joyeecheung): when we snapshot the bootstrapped context,
// the inspector and diagnostics setup should after after deserialization.
#if HAVE_INSPECTOR && NODE_USE_V8_PLATFORM
//TODO(jc-lab): If InitializeInspector fails, the caller must be informed of the return_code.
env->InitializeInspector(nullptr);
#endif
```
I thought about the above suggestions.
However, the above proposal does not pass the EnvironmentTest.MultipleEnvironmentsPerIsolate test.
https://github.com/nodejs/node/blob/1a552f614b4ec2b9faf0bf932ad4e6ef4fdcaf01/src/inspector_agent.cc#L770 | embedding | low | Critical |
546,743,004 | pytorch | error when building pytorch 1.1.0 from source | ## 🐛 Bug
When building pytorch 1.1.0 from source with explicitly setting `-D_GLIBCXX_USE_CXX11_ABI=0`, I met such error:
```
pytorch/build/lib/libcaffe2_gpu.so: undefined reference to `gloo::getCudaPCIBusID(int)'
pytorch/build/lib/libcaffe2_gpu.so: undefined reference to `gloo::EnforceNotMet::EnforceNotMet(char const*, int, char const*, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&)'
```
how to solve that?
**Without `export CFLAGS="-D_GLIBCXX_USE_CXX11_ABI=0 $CFLAGS"`**, I can build pytorch successfully, but when importing extend module which is built also from source in the same environment from python , I will meet error`my_module.so: undefined symbol: _ZN3c105ErrorC1ENS_14SourceLocationERKSs`, and my `torch._C._GLIBCXX_USE_CXX11_ABI == True`.(referenced [here](https://discuss.pytorch.org/t/undefined-symbol-when-import-lltm-cpp-extension/32627))
## To Reproduce
Steps to reproduce the behavior:
```
git clone https://github.com/pytorch/pytorch.git && git checkout v1.1.0
... # do some change to reset the third_party submodules to the correct commit with tags/v1.1.0
export USE_NINJA=OFF
export CFLAGS="-D_GLIBCXX_USE_CXX11_ABI=0 $CFLAGS"
export CMAKE_PREFIX_PATH=${CONDA_PREFIX:-"$(dirname $(which conda))/../"}
python setup.py install
```
<!-- If you have a code sample, error messages, stack traces, please provide it here as well -->
## Expected behavior
<!-- A clear and concise description of what you expected to happen. -->
## Environment
- PyTorch Version (e.g., 1.0): 1.1.0
- OS (e.g., Linux): Linux version 3.10.0-327.ali2012.alios7.x86_64
- How you installed PyTorch (`conda`, `pip`, source): source
- Build command you used (if compiling from source):
```
export USE_NINJA=OFF
export CFLAGS="-D_GLIBCXX_USE_CXX11_ABI=0 $CFLAGS"
export CMAKE_PREFIX_PATH=${CONDA_PREFIX:-"$(dirname $(which conda))/../"}
python setup.py install
```
- Python version: 3.6.9 | Anaconda [GCC 7.3.0] on linux
- CUDA/cuDNN version: cuda10.0
- GPU models and configuration: nvidia tesla v100 16GB
- gcc: 5.5.0 | module: build,triaged | low | Critical |
546,777,198 | flutter | Window geometry is incorrect after switching displays | When a flutter app gets moved from a hidpi display to a normal one, the window geometry can get messed up so that the window contents are offset vertically by the difference between the window height, and twice the window height. Or something like that - it's weird.
Here is a video that makes the issue clear:
[flutter_desktop_weirdness.mov.zip](https://github.com/flutter/flutter/files/4034795/flutter_desktop_weirdness.mov.zip)
This was originally reported as part of #37676 which was closed - the crash reported in that bug appears to have been fixed, but the broken geometry has not.
## Steps to Reproduce
1. Create a flutter desktop app.
2. Run it on a Mac laptop with no displays connected.
3. Close the laptop lid.
4. Connect it to a normal external display.
I'm not sure if this happens 100% of the time.
## Logs
There aren't any relevant exceptions or analysis errors.
```
[✓] Flutter (Channel master, v1.13.7-pre.50, on Mac OS X 10.14.4 18E226, locale en-GB)
• Flutter version 1.13.7-pre.50 at /Users/timh/local/flutter
• Framework revision 134c2ff076 (14 hours ago), 2020-01-07 12:23:02 -0800
• Engine revision 3851981b86
• Dart version 2.8.0 (build 2.8.0-dev.0.0 2f57602411)
[✓] Android toolchain - develop for Android devices (Android SDK version 28.0.3)
• Android SDK at /Users/timh/Library/Android/sdk
• Android NDK location not configured (optional; useful for native profiling support)
• Platform android-28, build-tools 28.0.3
• Java binary at: /Library/Java/JavaVirtualMachines/jdk1.8.0_152.jdk/Contents/Home/bin/java
• Java version Java(TM) SE Runtime Environment (build 1.8.0_152-b16)
• All Android licenses accepted.
[✓] Xcode - develop for iOS and macOS (Xcode 11.3)
• Xcode at /Applications/Xcode.app/Contents/Developer
• Xcode 11.3, Build version 11C29
• CocoaPods version 1.7.1
[!] Android Studio (not installed)
• Android Studio not found; download from https://developer.android.com/studio/index.html
(or visit https://flutter.dev/setup/#android-setup for detailed instructions).
[✓] VS Code (version 1.41.1)
• VS Code at /Applications/Visual Studio Code.app/Contents
• Flutter extension version 3.7.1
[✓] Connected device (1 available)
• macOS • macOS • darwin-x64 • Mac OS X 10.14.4 18E226
! Doctor found issues in 1 category.
```
| e: device-specific,framework,engine,platform-mac,c: rendering,a: desktop,P2,team-macos,triaged-macos | low | Critical |
546,781,113 | godot | ScollContainer's ScrollBar doesn't interact other than mouse. | **Godot version:**
3.1.2
**OS/device including version:**
Windows 10
**Issue description:**
VScrollBarr doesn't get focus when the button inside ScrollContainer/VBoxContainer is focused and action "ui_right" is pressed.
ScrollContainer.get_v_scrollbar( ).value and max_value doesn't represent position or range.
Setting value on VScrollBar doesn't position it correctly - uncertain behavior. Even giving it the max_value it doesn't go past the middle.
There's no other way to position VScrollBar with code. At least the documentation doesn't have ANYTHING about ScrollBar interaction.
What action is triggered when the mouse wheel is moved and scrolls the "page"? There's no action in InputMap that uses a mouse wheel.
P.S.
There's no documentation on how to use separate ScrollBar (where to put and how to interact with it).
**Steps to reproduce:**
Just VBoxContainer that extends more than ScrollContainer allows, so VScrollBarr appears.
**Minimal reproduction project:**
[ScrollContainerBug.zip](https://github.com/godotengine/godot/files/4034826/ScrollContainerBug.zip)
| bug,topic:core,confirmed | low | Critical |
546,783,241 | flutter | driver.tap based on x,y co-ordinates | How to do driver.tap based on x,y coordinates on the device | a: tests,c: new feature,framework,d: api docs,t: flutter driver,P3,team-framework,triaged-framework | low | Major |
546,796,831 | godot | GraphEdit doesn't use the closest ports when dragging | **Godot version:**
8b5992f6658a125f63424a687e15e75ccd36906c
**Issue description:**
When dragging out from a GraphNode it doesn't use the closest port.
**Steps to reproduce:**
Put two GraphNode ports close to another and try dragging out from them until the issue happens.
This can be done in VisualScript.
**Code responsible:**
https://github.com/godotengine/godot/blob/8b5992f6658a125f63424a687e15e75ccd36906c/scene/gui/graph_edit.cpp#L594-L621
When checking if a port is clicked, it just loops over the ports and finds the first one where `is_is_hot_zone` is true. | bug,confirmed,usability,topic:gui | low | Minor |
546,829,224 | opencv | dnn(permute): reduce permute to reshape (or transpose) whenever possible and work inplace | ##### System information (version)
- OpenCV => 4.2.0
- Operating System / Platform => Ubuntu 18.04
- Compiler => GCC 7.4
##### Detailed description
Vast majority of permute operations can be reduced to a reshape or a 2d-transpose. The reshape operation can be skipped entirely and the transpose operation can be efficiently performed inplace.
**Stats for single image inference:**
Model | Total Permute Layers | Reduced to 2d-transpose | Reduced to reshape
---------------------- | ---------------------------- | ----------------------------- | ---------------------------
MobileNet SSD | 12 | 10 | 2
YOLOv3 | 3 | 3 | 0
Inception v2 Faster RCNN | 3 | 2 | 1
Inception v2 Mask RCNN | 3 | 2 | 1
Currently, the permute layer operates inplace when the permute order is identity order. This logic can be upgraded to include the logic used here:
1. https://github.com/opencv/opencv/blob/1f2b2c52422813188bc5cbcc5d312c89e41786b2/modules/dnn/src/cuda/permute.cu#L173-L186
2. https://github.com/opencv/opencv/blob/1f2b2c52422813188bc5cbcc5d312c89e41786b2/modules/dnn/src/cuda/permute.cu#L217-L231
This logic would allow the permute operation to be skipped entirely when it reduces to a reshape and perform inplace when it reduces to a 2d-transpose.
The reshape optimization is easier to implement. The logic can be implemented in `getMemoryShapes`. The same logic can be implemented in `op_permute.cpp` for the Vulkan backend. It's fairly easy to implement the same in the CUDA backend. IE nodes created in `BlankLayer` can be used when permute reduces to a reshape (or maybe IE's PermuteLayer/TransposeOp could handle it on its own).
The inplace 2d-transpose optimization (as all backends may not support it) would require inplace to be triggered for specific backends (which I believe the current DNN module doesn't support). | optimization,feature,category: dnn | low | Major |
546,877,221 | rust | Add ability to ignore tests at runtime. | For medium–large scale projects, or projects that have a lot of integrations with third party services. It would nice to be able to ignore tests based on the environment the tests are being run in. One example would be to run tests if an environment variable is present.
You can workaround this with something like the following, however the output of the test is `ok` and not `ignored`, misleading the user that the test was successful.
```rust
#[test]
fn foo() {
if env::var("FOO").is_none() { return; }
let foo = env::var("FOO").unwrap();
}
``` | T-lang,T-dev-tools,C-feature-request,A-libtest | medium | Critical |
546,957,258 | pytorch | Function request: logerfc, logerfcx special functions | ## 🚀 Feature
Implement the `erfcx(x)` special function, which computes `exp(x^2) * erfc(x)` in a numerically stable way. Also for convenience, add `logerfc(x) = log(erfc(x))` and `logerfcx(x) = log(erfcx(x))`.
`erfcx` is available in many numerical packages, such as [Matlab](https://www.mathworks.com/help/matlab/ref/erfcx.html), [Julia](https://juliamath.github.io/SpecialFunctions.jl/latest/functions_list/#SpecialFunctions.erfcx), [SciPy](https://docs.scipy.org/doc/scipy/reference/generated/scipy.special.erfcx.html) [R](https://www.rdocumentation.org/packages/pracma/versions/1.9.9/topics/erf), and others.
From `erfcx` it is easy to obtain `logerfc` and `logerfcx`, but this involves a conditional which can be slow in pure Python code. So I recommend adding `logerfc` and `logerfcx` as well, which can be implemented as:
```
def logerfc(x):
if x > 0.0:
return log(erfcx(x)) - x**2
else:
return log(erfc(x))
def logerfcx(x):
if x < 0.0:
return log(erfc(x)) + x^2
else:
return log(erfcx(x))
```
## Motivation
These special functions are very useful whenever we have to work with truncated normal distributions.
Related: https://github.com/pytorch/pytorch/issues/2129, https://github.com/pytorch/pytorch/issues/32293
cc @mruberry @rgommers @heitorschueroff | module: numerical-stability,triaged,module: numpy,function request,module: special | low | Major |
546,976,888 | flutter | [web] Canvas.saveLayer does not work in HTML mode | ```dart
canvas.drawRect(...);
canvas.saveLayer(null, Paint()..blendMode = BlendMode.multiply);
canvas.drawRect(...);
canvas.restore();
```
This works fine when running Flutter on Android and looks like this:
[![Screen capture][1]][1]
When running the same code in Flutter Web, it looks like this (`saveLayer` is ignored):
[![Screen capture][2]][2]
When I move the `blendMode` to the `Paint` I pass to the second `drawRect`, it also works on Flutter web.
```
flutter doctor -v
[√] Flutter (Channel master, v1.13.7-pre.41, locale en-US)
• Flutter version 1.13.7-pre.41 at flutter
• Framework revision fa337f5922 (2 days ago), 2020-01-06 16:44:12 -0800
• Engine revision 3851981b86
• Dart version 2.8.0 (build 2.8.0-dev.0.0 2f57602411)
```
[2]: https://i.stack.imgur.com/q3In0.png
[1]: https://i.stack.imgur.com/s84vd.png | engine,platform-web,c: rendering,e: web_html,has reproducible steps,P3,team-web,triaged-web,found in release: 3.13,found in release: 3.17 | medium | Major |
546,985,671 | rust | `Pin` is unsound due to transitive effects of `CoerceUnsized` | Split out from #66544. It is possible to exploit `Pin` on nightly Rust (but not stable) by creating smart pointers that implement `CoerceUnsized` but have strange behavior. See [the dedicated internals thread](https://internals.rust-lang.org/t/the-pin-unsoundness-and-coerceunsized/11593) for more details -- also, please keep conversation on the thread, and not on the Github issue. ❤️
| P-medium,T-lang,I-unsound,C-bug,E-needs-mcve,requires-nightly,F-coerce_unsized | low | Major |
547,010,743 | godot | Shadows are bugged with tiles in tilemaps | **Godot version:** 3.1.2 stable
**OS:** RebornOS (kernel version: Linux 5.4.2-arch1-1 x86_64)
**Issue description:** Look at picture
I tried to google it but nothing
**Steps to reproduce:** Create 2 scenes, one with tiles, one the main scene
In the tile scene create some tiles and add LightOccluder2D to it and add CounterClockwise
In the main scene add Tilemap with tiles from the previous scene, add something to move with Light2D and enable shadows and use as light the image below
| bug,topic:rendering,topic:2d | low | Critical |
547,034,115 | pytorch | JIT tests are linked directly into libtorch and register operators even when unused | The two custom classes defined in [test/cpp/jit/test_custom_class.cpp](https://github.com/pytorch/pytorch/blob/7f723cbd8a53c85722af14d8749559de1992cd9b/test/cpp/jit/test_custom_class.cpp#L49-L61) will register new jit operators as soon as the library is loaded.
This lead to different operators being registered if we build the tests or not and that trips the backward compatibility checks.
https://github.com/pytorch/pytorch/pull/31949 is a temporary fix but should be improved.
The possible approaches here are:
- move the cpp tests outside of libtorch to avoid this issue altogether (and that will reduce release binary size!)
- change the test to only register these operators if they are actually used
- permanently whitelist all the `_TorchScriptTesting*` for the backward compatibility checks and ignore this issue.
cc @suo | oncall: jit,triaged | low | Minor |
547,034,622 | go | proposal: text/template: revise execution model | Go's template packages (`text/template`, `html/template`) do in general work really well. But there are some issues that are hard/impossible to work around.
In the current implementation, a `Template` struct is mirrored in both packages, and while there is some separation of parsing and execution to allow for concurrent execution, it is very much intermingled.
This proposal suggests adding a new "template executer" which would:
1. Prepare and execute the template
2. Store and resolve any template functions
3. Provide some user-defined hook interface to allow for a custom evaluation of method/func/map lookups etc.
The above may look like different _things_, candidates for different _proposals_. But these issues are connected and I'm convinced that it helps to look at them as a whole, even if this proposal is only partly accepted.
The benefits of the above would be:
2. Stateful functions. One common example would be a `i18n` translation func. You could make it a method (`{{ .I18n "hello" }}`) or make it take the language as the first argument (`{{ i18n .Language "hello" }}`), neither very practical. The current workaround is to clone the template set for each language, which is both resource wasteful and hard to get right in the non-trivial setups. I assume that these functions are added early to get parse-time signature validation, a good thing, but I don't see a reason why not to use another set for execution, making the function lookup lock-free a great performance bonus.
3. Some examples include case insensitive map lookups, adding some execution context to methods that support it (https://github.com/golang/go/issues/31107), allow range of funcs that return map/slice (https://github.com/golang/go/issues/20503), allow a custom variant of the built-in and rather opinionated version of `IsTrue` (https://github.com/golang/go/issues/28391)
Doing this also removes some performance bottlenecks. A relevant benchmark in Hugo before/after we made these changes:
```bash
name old time/op new time/op delta
SiteNew/Many_HTML_templates-16 55.6ms ± 2% 42.9ms ± 1% -22.81% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
SiteNew/Many_HTML_templates-16 22.5MB ± 0% 17.6MB ± 0% -21.99% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
SiteNew/Many_HTML_templates-16 341k ± 0% 247k ± 0% -27.48% (p=0.008 n=5+5)
```
## Implementation outline
In the [Hugo project](https://github.com/gohugoio/hugo) we have worked around the problems outlined above, some workaround less pretty than others. But recently we stumbled into a hurdle we didn't find a way around, so we closed our eyes and created a scripted fork of the `text/template` and `html/template` packages. We will pull in upstream fixes, and the ultimate goal is for that fork to go away.
But this means that there is a working implementation of this proposal. The implementation is a little bit coloured by trying to do the least amount of changes to the existing code, but I think it outlines a possible API. And it is backwards compatible.
```go
// Executer executes a given template.
type Executer interface {
Execute(p Preparer, wr io.Writer, data interface{}) error
}
```
```go
// Preparer prepares the template before execution.
type Preparer interface {
Prepare() (*Template, error)
}
```
```go
// ExecHelper allows some custom eval hooks.
type ExecHelper interface {
GetFunc(tmpl Preparer, name string) (reflect.Value, bool)
GetMethod(tmpl Preparer, receiver reflect.Value, name string) (method reflect.Value, firstArg reflect.Value)
GetMapValue(tmpl Preparer, receiver, key reflect.Value) (reflect.Value, bool)
IsTrue(val reflect.Value) bool
}
```
Hugo's fork can be found here https://github.com/gohugoio/hugo/tree/master/tpl/internal/go_templates (all the patches are placed in `hugo_*.go` files). Changes are marked with the comment `Added for Hugo.`. The script that maintains the fork can be found at https://github.com/gohugoio/hugo/tree/master/scripts/fork_go_templates.
## Some related Go issues
* Add ExecuteContext methods https://github.com/golang/go/issues/31107
* Document ability to modify FuncMap after template parse by calling Funcs again https://github.com/golang/go/issues/34680
* Allow callers to override IsTrue https://github.com/golang/go/issues/28391
* Methods vs funcs discrepancy https://github.com/golang/go/issues/20503
* index should return nil instead of index out of range error https://github.com/golang/go/issues/14751 | v2,Proposal | medium | Critical |
547,041,392 | PowerToys | touch/update timestamp for Windows-GUI in file-context-menu | PowerToy-suggestion
scenario:
I move a bunch of files from folder A to folder B (which contains several hundred files).
From my perspective they're "new" in folder B. And ordering my files by "date modified" is an important part of my organization.
However "date modified" does not update, if files or folders merely change location, so it fails at sorting by "new". If I add in twenty new files and they're all at twenty random wrong places in my long list.
However when I just move them in, all of them will still be selected. It would be really convenient, if I could right-click one of the selected files and click "update 'last modified'-timestamp".
why I need that:
I use a graph-based organization tool called "TheBrain 11" to manage all my files, instead of the hierarchical tree-structure. So all my personal files actually "live" in one folder. Each file is linked to a Thought in TheBrain (or to several). However there's still files I haven't added to that folder yet, partly because it's inconvenient to do so, because I can't differentiate in the file-explorer of whether they're in TheBrain already or not.
If there's a tool already out there, that does exactly that, I'd be thankful for any pointers.
Also, this would be good to have for folders as well. | Idea-New PowerToy | low | Major |
547,046,636 | flutter | Consider not setting overflow:hidden on text with no max lines | We currently set `overflow:hidden` on all `<p>`, but only those with maxlines need it. In general, clipping is considered an expensive operation, so avoiding it could have performance benefits. It's hard to tell how much cost this incurs in our case, but filing it anyway, so we can measure and fix it, or just blindly fix it on a hunch if performance investigation is not worth the trouble. | engine,c: performance,platform-web,c: proposal,perf: speed,e: web_html,P3,team-web,triaged-web | low | Major |
547,059,765 | kubernetes | Improve CRD openapi serving speed | CustomResourcePublishOpenAPI e2e tests are [flaky](https://github.com/kubernetes/kubernetes/issues/86418) because the time for a given CRD change to get reflected in the openapi spec exceeds the 30s timeout.
Here is a analysis for the worst case-- assuming three CRD changes were made at the same time (change A&B by tests running in parallel, and change C by the current test), it may take up to three 200OK downloads before the current e2e test sees change C in the openapi spec. The 30s timeout can be reached before the test starts the third 200OK download.
Things that we can improve:
1. [openapi service update](https://github.com/kubernetes/kubernetes/blob/b34c96b62c74c22ac9ac135fc787016786845a69/vendor/k8s.io/kube-openapi/pkg/handler/handler.go#L106-L139) (for both kube-aggregator and CRD) is most expensive. [Adding log](https://github.com/kubernetes/kubernetes/pull/86564) shows both the first [marshaling](https://github.com/kubernetes/kubernetes/blob/b34c96b62c74c22ac9ac135fc787016786845a69/vendor/k8s.io/kube-openapi/pkg/handler/handler.go#L107) and [ToProtoBinary](https://github.com/kubernetes/kubernetes/blob/b34c96b62c74c22ac9ac135fc787016786845a69/vendor/k8s.io/kube-openapi/pkg/handler/handler.go#L115) can cost more that 5s for kube-aggregator. The time spent also varies a lot even though the spec size should be rather stable (due to CPU pressure?). More efficient ways to do the marshaling and generate proto binary would help.
2. although the CRD openapi spec is much smaller compared to kube-aggregator, openapi service update still costs 0.5~1.5s to apiextensions-apiserver for each CRD change. Since openapi service update requires lock, the time queues up when there are lots of CRD changes at the same time (e.g. three CRD changes in the image above). Solving the previous point also helps here
3. apiextensions-apiserver supports serving openapi spec in PB and gzip, but kube-aggregator downloads and deserializes the spec through JSON (which makes sense for aggregated apiservers). Potentially kube-aggregator can use PB when it's talking to apiextensions-apiserver.
cc @liggitt @sttts @wojtek-t | priority/important-soon,sig/scalability,sig/api-machinery,kind/feature,area/custom-resources,lifecycle/frozen | low | Major |
547,078,288 | godot | Array out of bounds in libtheora | **Godot version:**
Godot 3.2 beta 5
**OS/device including version:**
Ubuntu 19.10
**Issue description:**
Backtrace
```
thirdparty/libtheora/state.c:327:37: runtime error: left shift of 1 by 63 places cannot be represented in type 'long int'
thirdparty/libtheora/state.c:576:19: runtime error: left shift of negative value -800
thirdparty/libtheora/huffdec.c:211:43: runtime error: index 4 out of bounds for type 'oc_huff_node *[2]'
thirdparty/libtheora/huffdec.c:210:56: runtime error: index 4 out of bounds for type 'oc_huff_node *[2]'
thirdparty/libtheora/huffdec.c:329:33: runtime error: index 2 out of bounds for type 'oc_huff_node *[2]'
thirdparty/libtheora/huffdec.c:329:47: runtime error: index 2 out of bounds for type 'oc_huff_node *[2]'
thirdparty/libtheora/huffdec.c:327:51: runtime error: index 4 out of bounds for type 'oc_huff_node *[2]'
thirdparty/libtheora/huffdec.c:327:17: runtime error: index 4 out of bounds for type 'oc_huff_node *[2]'
thirdparty/libtheora/huffdec.c:328:42: runtime error: index 4 out of bounds for type 'oc_huff_node *[2]'
```
**Steps to reproduce:**
1. Compile Godot from https://github.com/qarmin/godot/tree/bb, or apply patch from this commit to your own branch with GCC 9 and Clang 9
2. Run project in editor
**Minimal reproduction project:**
https://github.com/GDQuest/godot-power-pitch | bug,topic:thirdparty | low | Critical |
547,105,280 | TypeScript | Generator prototype ignored | **TypeScript Version:** 3.8.0-dev.20200107
**Search Terms:** generator, iterator, prototype, inherit
**Code**
```ts
function* gen() {
while (true) {
yield "something";
}
}
gen.prototype.test = () => {
console.log("OK");
};
gen().test();
```
**Expected behavior:** No errors
**Actual behavior:** Error `TS2339: Property 'test' does not exist on type 'Generator<string, void, unknown>'`
**Playground Link:** <!-- A link to a TypeScript Playground "Share" link which demonstrates this behavior --> [Provided](https://www.typescriptlang.org/play/?target=99&ts=3.8.0-dev.20200107#code/GYVwdgxgLglg9mAVAAgOYFMwAoCUyDeAsAFDJnIDuAFjADbrJZQBOI6eRp53AnjOrQAmyAEQBnOAFt0UGmFQiA3CW4BfEuuIkMYAHQAHZnCjGe+9LqjoxUZAF5GeOwD4CK8hAQT6u2nFRYIgDyANIiOMrEqpHamLiW1lC4isgA9KnIAKLMRszIACoAygBMAMylAJwAXMgACkbmzFA8yADkVjatyIJw1shgxsjoAB4wNsgIyM3mbQDimOjMAIYmzAA8Nsww8gA0yABucDCCe+AA1gMUYM6tuipAA)
| Suggestion,Awaiting More Feedback | low | Critical |
547,108,971 | terminal | Remove usages of old safe math | There are approximately 79 uses of old safe math. (a.k.a. the `LongAdd`, `ShortAdd`, and friends from `intsafe.h`).
This task represents removing them and transitioning them to an appropriate safe/saturating math function from the chromium safe math library introduced in #4144.
After removing usages of the old safe math functions, also remove the `intsafe.h` header.
Files with old safe math as of this writing:
- TerminalControl: TSFInputControl.cpp
- TerminalCore: TerminalDispatch.cpp
- TerminalCore: TerminalSelection.cpp
- Host: directio.cpp
- RendererGdi: invalidate.cpp
- RendererGdi: math.cpp
- RendererGdi: paint.cpp
- RendererVt: paint.cpp
- TerminalAdapter: adaptDispatch.cpp
- Types: viewport.cpp
- Types: WindowUiaProviderBase.cpp | Product-Conhost,Help Wanted,Product-Conpty,Issue-Task,Area-CodeHealth | low | Minor |
547,122,006 | kubernetes | Remove VerifyVolumeAttached for CSI Plugin | /cleanup
/assign
VerifyVolumeAttach functionality for CSI is handled by default by the [csi-attacher v2.1+](https://github.com/kubernetes-csi/external-attacher/blob/v2.1.0/CHANGELOG-2.1.md). The in-tree verification is a no-op since it just looks at the `VolumeAttachment.Status.Attached` field which is not an accurate representation of reality when the backend attach status has changed out of band.
/cc @msau42 @saad-ali | sig/storage,kind/feature,lifecycle/frozen | low | Minor |
547,124,336 | flutter | InkWell.onLongPress cancels ripple animation | I am trying to build a button that increments a number when the user taps on it, then continues to increment that number every `x` milliseconds while the user continues to long press the button.
If I try out this code sample, it behaves as expected. The number is incremented one time, and the ripple effect works properly:
```dart
return Material(
child: InkWell(
onTap: () { increment(); },
child: Container(
width: 30,
height: 30,
decoration: BoxDecoration(
border: Border.all(color: Colors.black),
),
),
),
);
```
If I then try to implement the `onLongPress` callback, even just an empty function will stop the ripple effect immediately:
```dart
return Material(
child: InkWell(
onTap: () { increment(); },
onLongPress: () {}, // this breaks the ripple effect
child: Container(
width: 30,
height: 30,
decoration: BoxDecoration(
border: Border.all(color: Colors.black),
),
),
),
);
```
There doesn't seem to be a way to use the `onLongPress` parameter without breaking the ripple effect. | framework,f: material design,f: gestures,customer: crowd,has reproducible steps,found in release: 3.3,customer: flex,found in release: 3.7,team-design,triaged-design | low | Major |
547,142,101 | go | crypto/tls: Client Side TLS Authentication fails for certs with long fields | ### What version of Go are you using (`go version`)?
<pre>
go version go1.13.4 linux/amd64
</pre>
### Does this issue reproduce with the latest release?
Yes
### What operating system and processor architecture are you using (`go env`)?
<details><summary><code>go env</code> Output</summary><br><pre>
GO111MODULE=""
GOARCH="amd64"
GOBIN=""
GOCACHE="/home/user/.cache/go-build"
GOENV="/home/user/.config/go/env"
GOEXE=""
GOFLAGS=""
GOHOSTARCH="amd64"
GOHOSTOS="linux"
GONOPROXY=""
GONOSUMDB=""
GOOS="linux"
GOPATH="/home/user/Revisions/go"
GOPRIVATE=""
GOPROXY="https://proxy.golang.org,direct"
GOROOT="/usr/lib/go-1.13"
GOSUMDB="sum.golang.org"
GOTMPDIR=""
GOTOOLDIR="/usr/lib/go-1.13/pkg/tool/linux_amd64"
GCCGO="gccgo"
AR="ar"
CC="gcc"
CXX="g++"
CGO_ENABLED="1"
GOMOD="/home/user/go/src/helm.sh/helm/go.mod"
CGO_CFLAGS="-g -O2"
CGO_CPPFLAGS=""
CGO_CXXFLAGS="-g -O2"
CGO_FFLAGS="-g -O2"
CGO_LDFLAGS="-g -O2"
PKG_CONFIG="pkg-config"
GOGCCFLAGS="-fPIC -m64 -pthread -fmessage-length=0 -fdebug-prefix-map=/tmp/go-build386574731=/tmp/go-build -gno-record-gcc-switches"
</pre></details>
### What did you do?
package main
import (
"crypto/tls"
"fmt"
)
func main() {
cert, err := tls.LoadX509KeyPair("./helm-test-crt", "./helm-test.key")
fmt.Println(cert, err)
}
Test certs:
https://gist.github.com/antevens/a05409165d33bd771e39bc219fe2c6bb/archive/e56cf114dee64a6b367abf08ab1c34a405a62f88.zip
https://github.com/helm/helm/issues/7343
### What did you expect to see?
Client side TLS authentication succeed.
### What did you see instead?
asn1: structure error: base 128 integer too large | NeedsInvestigation | low | Critical |
547,174,150 | go | x/playground: out of memory crashing | I was looking at the logs for the playground and see lots of crashes like:
```
A 2020-01-08T22:59:11Z fatal error: runtime: out of memory
A 2020-01-08T22:59:11Z
A 2020-01-08T22:59:11Z runtime stack:
A 2020-01-08T22:59:11Z runtime.throw(0xbc60ed, 0x16)
A 2020-01-08T22:59:11Z /usr/local/go/src/runtime/panic.go:774 +0x72
A 2020-01-08T22:59:11Z runtime.sysMap(0xc088000000, 0x24000000, 0x12028f8)
A 2020-01-08T22:59:11Z /usr/local/go/src/runtime/mem_linux.go:169 +0xc5
A 2020-01-08T22:59:11Z runtime.(*mheap).sysAlloc(0x11ea160, 0x20040000, 0x7f5eb32eed08, 0x4377c7)
A 2020-01-08T22:59:11Z /usr/local/go/src/runtime/malloc.go:701 +0x1cd
A 2020-01-08T22:59:11Z runtime.(*mheap).grow(0x11ea160, 0x10020, 0xffffffff)
A 2020-01-08T22:59:11Z /usr/local/go/src/runtime/mheap.go:1255 +0xa3
A 2020-01-08T22:59:11Z runtime.(*mheap).allocSpanLocked(0x11ea160, 0x10020, 0x1202908, 0x1195e80)
A 2020-01-08T22:59:11Z /usr/local/go/src/runtime/mheap.go:1170 +0x266
A 2020-01-08T22:59:11Z runtime.(*mheap).alloc_m(0x11ea160, 0x10020, 0x101, 0x0)
A 2020-01-08T22:59:11Z /usr/local/go/src/runtime/mheap.go:1022 +0xc2
A 2020-01-08T22:59:11Z runtime.(*mheap).alloc.func1()
A 2020-01-08T22:59:11Z /usr/local/go/src/runtime/mheap.go:1093 +0x4c
A 2020-01-08T22:59:11Z runtime.(*mheap).alloc(0x11ea160, 0x10020, 0x7f5eb3000101, 0x429510)
A 2020-01-08T22:59:11Z /usr/local/go/src/runtime/mheap.go:1092 +0x8a
A 2020-01-08T22:59:11Z runtime.largeAlloc(0x20040000, 0x450100, 0xc04be66000)
A 2020-01-08T22:59:11Z /usr/local/go/src/runtime/malloc.go:1138 +0x97
A 2020-01-08T22:59:11Z runtime.mallocgc.func1()
A 2020-01-08T22:59:11Z /usr/local/go/src/runtime/malloc.go:1033 +0x46
A 2020-01-08T22:59:11Z runtime.systemstack(0x0)
A 2020-01-08T22:59:11Z /usr/local/go/src/runtime/asm_amd64.s:370 +0x66
A 2020-01-08T22:59:11Z runtime.mstart()
A 2020-01-08T22:59:11Z /usr/local/go/src/runtime/proc.go:1146
A 2020-01-08T22:59:11Z
A 2020-01-08T22:59:11Z goroutine 416 [running]:
A 2020-01-08T22:59:11Z runtime.systemstack_switch()
A 2020-01-08T22:59:11Z /usr/local/go/src/runtime/asm_amd64.s:330 fp=0xc000157058 sp=0xc000157050 pc=0x45a6b0
A 2020-01-08T22:59:11Z runtime.mallocgc(0x20040000, 0x0, 0x0, 0x0)
A 2020-01-08T22:59:11Z /usr/local/go/src/runtime/malloc.go:1032 +0x895 fp=0xc0001570f8 sp=0xc000157058 pc=0x40e1b5
A 2020-01-08T22:59:11Z runtime.slicebytetostring(0x0, 0xc04be66000, 0x20040000, 0x20800bcc, 0x0, 0x0)
A 2020-01-08T22:59:11Z /usr/local/go/src/runtime/string.go:102 +0x9f fp=0xc000157128 sp=0xc0001570f8 pc=0x448dff
A 2020-01-08T22:59:11Z main.(*Recorder).Events(0xc00039e8c0, 0x0, 0x0, 0x6, 0x6, 0xc0003022c0)
A 2020-01-08T22:59:11Z /go/src/playground/play.go:93 +0x36f fp=0xc0001572e0 sp=0xc000157128 pc=0x9fe9ef
A 2020-01-08T22:59:11Z main.compileAndRun(0xd35260, 0xc00026ec80, 0xc0003869a0, 0x0, 0x0, 0x0)
A 2020-01-08T22:59:11Z /go/src/playground/sandbox.go:493 +0x14ed fp=0xc000157940 sp=0xc0001572e0 pc=0xa0220d
A 2020-01-08T22:59:11Z main.(*server).commandHandler.func1(0xd336e0, 0xc0000cea80, 0xc0002f2800)
A 2020-01-08T22:59:11Z /go/src/playground/sandbox.go:115 +0x468 fp=0xc000157b30 sp=0xc000157940 pc=0xa09df8
A 2020-01-08T22:59:11Z net/http.HandlerFunc.ServeHTTP(0xc0002734a0, 0xd336e0, 0xc0000cea80, 0xc0002f2800)
A 2020-01-08T22:59:11Z /usr/local/go/src/net/http/server.go:2007 +0x44 fp=0xc000157b58 sp=0xc000157b30 pc=0x753d34
A 2020-01-08T22:59:11Z net/http.(*ServeMux).ServeHTTP(0xc00026eac0, 0xd336e0, 0xc0000cea80, 0xc0002f2800)
A 2020-01-08T22:59:11Z /usr/local/go/src/net/http/server.go:2387 +0x1bd fp=0xc000157bb8 sp=0xc000157b58 pc=0x755d0d
A 2020-01-08T22:59:11Z main.(*server).ServeHTTP(0xc0000ebdb0, 0xd336e0, 0xc0000cea80, 0xc0002f2800)
A 2020-01-08T22:59:11Z /go/src/playground/server.go:92 +0x193 fp=0xc000157c18 sp=0xc000157bb8 pc=0xa05043
A 2020-01-08T22:59:11Z net/http.serverHandler.ServeHTTP(0xc000362000, 0xd336e0, 0xc0000cea80, 0xc0002f2800)
A 2020-01-08T22:59:11Z /usr/local/go/src/net/http/server.go:2802 +0xa4 fp=0xc000157c48 sp=0xc000157c18 pc=0x757284
A 2020-01-08T22:59:11Z net/http.(*conn).serve(0xc00019c140, 0xd35260, 0xc00007a040)
A 2020-01-08T22:59:11Z /usr/local/go/src/net/http/server.go:1890 +0x875 fp=0xc000157fc8 sp=0xc000157c48 pc=0x752b25
A 2020-01-08T22:59:11Z runtime.goexit()
A 2020-01-08T22:59:11Z /usr/local/go/src/runtime/asm_amd64.s:1357 +0x1 fp=0xc000157fd0 sp=0xc000157fc8 pc=0x45c781
A 2020-01-08T22:59:11Z created by net/http.(*Server).Serve
A 2020-01-08T22:59:11Z /usr/local/go/src/net/http/server.go:2928 +0x384
A 2020-01-08T22:59:11Z
A 2020-01-08T22:59:11Z goroutine 1 [IO wait]:
A 2020-01-08T22:59:11Z internal/poll.runtime_pollWait(0x7f5eb6446fe8, 0x72, 0x0)
A 2020-01-08T22:59:11Z /usr/local/go/src/runtime/netpoll.go:184 +0x55
A 2020-01-08T22:59:11Z internal/poll.(*pollDesc).wait(0xc0002e2898, 0x72, 0x0, 0x0, 0xba4397)
A 2020-01-08T22:59:11Z /usr/local/go/src/internal/poll/fd_poll_runtime.go:87 +0x45
A 2020-01-08T22:59:11Z internal/poll.(*pollDesc).waitRead(...)
A 2020-01-08T22:59:11Z /usr/local/go/src/internal/poll/fd_poll_runtime.go:92
A 2020-01-08T22:59:11Z internal/poll.(*FD).Accept(0xc0002e2880, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0)
A 2020-01-08T22:59:11Z /usr/local/go/src/internal/poll/fd_unix.go:384 +0x1f8
A 2020-01-08T22:59:11Z net.(*netFD).accept(0xc0002e2880, 0xc0005c9c80, 0x758214, 0xc0003620a0)
A 2020-01-08T22:59:11Z /usr/local/go/src/net/fd_unix.go:238 +0x42
A 2020-01-08T22:59:11Z net.(*TCPListener).accept(0xc00000e4c0, 0x5e165eba, 0xc0005c9c80, 0x4c90a6)
A 2020-01-08T22:59:11Z /usr/local/go/src/net/tcpsock_posix.go:139 +0x32
A 2020-01-08T22:59:11Z net.(*TCPListener).Accept(0xc00000e4c0, 0xc0005c9cd0, 0x18, 0xc000000180, 0x757744)
A 2020-01-08T22:59:11Z /usr/local/go/src/net/tcpsock.go:261 +0x47
A 2020-01-08T22:59:11Z net/http.(*Server).Serve(0xc000362000, 0xd33420, 0xc00000e4c0, 0x0, 0x0)
A 2020-01-08T22:59:11Z /usr/local/go/src/net/http/server.go:2896 +0x280
A 2020-01-08T22:59:11Z net/http.(*Server).ListenAndServe(0xc000362000, 0xc000362000, 0x1)
A 2020-01-08T22:59:11Z /usr/local/go/src/net/http/server.go:2825 +0xb7
A 2020-01-08T22:59:11Z net/http.ListenAndServe(...)
A 2020-01-08T22:59:11Z /usr/local/go/src/net/http/server.go:3081
A 2020-01-08T22:59:11Z main.main()
A 2020-01-08T22:59:11Z /go/src/playground/main.go:58 +0x2cd
A 2020-01-08T22:59:11Z
```
I haven't looked into it yet but filing before I forget.
I don't think it's related to the new gvisor sandbox code (that was just submitted before I saw these logs) because the new code is disabled right now and I don't see how the other changes could cause this (too much memory being used by the Recorder type?).
Investigate.
| NeedsInvestigation | low | Critical |
547,191,654 | go | cmd/go: include source dir in cache key when ${SRCDIR} is present in cgo header | What I did:
* have a project including cgo headers using `${SRCDIR}`
* move that project to a new directory
* build
Result:
`ld: warning: directory not found for option '-L/Users/josh/...redacted...`
The redacted directory is the previous directory pointed to by the `${SRCDIR}` directive in the cgo sources.
After a `go clean -cache`, the error disappeared.
I suspect that cmd/go's caching may be too effective here. It probably doesn't take into account the absolute path containing the project, but it may need to when the `${SRCDIR}` cgo directive is present.
cc @bcmills @jayconrod
| NeedsFix,GoCommand | low | Critical |
547,200,280 | tensorflow | Define variable shape at restore/load, allow direct restoring of variables prior to calling __build__ (non-lazy variable loading from checkpoint) | <em>Please make sure that this is a feature request. As per our [GitHub Policy](https://github.com/tensorflow/tensorflow/blob/master/ISSUES.md), we only address code/doc bugs, performance issues, feature requests and build/installation issues on GitHub. tag:feature_template</em>
**System information**
- TensorFlow version (you are using): 2.0
- Are you willing to contribute it (Yes/No): Yes (if necessary)
**Describe the feature and the current behavior/state.**
Typically for RL, I use tensorflow in combination with Ray, where a remote agent collects data using a policy model, and the episodes are fed to a tf-agents EpisodicReplayBuffer in a second train-loop process. A tf.dataset is used to feed the main model training on this thread, which is then serialized via get_weights and sent back to the remote actor via a ray remote call to set_weights.
Currently I use tf.Checkpoint.restore.expect_partial() to restore my subclassed tf.keras.Model. In practice, this is extremely convoluted on initialization:
* I make a forward pass through my remote Agent on random policy (since I have not initialized the weights), retrieve this forward pass, and send it to the train loop
* I make a second forward pass through my train loop (again, randomly initialized network still), this has the side effect of initializing the network, allowing my checkpoint to restore the variables in this model
* I call get_weights on the train_loop model, and pass the weights to my remote model via set_weights.
* I can now collect a full replay buffer with my Agent on policy
* I can now run the train loop
Maybe I'm missing something, but most of my labmates are also confused by what the best practice for this currently is.
**Will this change the current api? How?**
I'm not sure what the best approach is. Some ideas for discussion?
* Allow set_weights to define the shapes and names of tf.Variables, such that calls to __build__ in the tf.keras.Model on first run will align with the initialized tf.Variables.
* Allow tf.keras.model.load_weights() to set the weights immediately rather than waiting for first call to build()
**Who will benefit with this feature?**
RL community
**Any Other info.**
| stat:awaiting tensorflower,type:feature,comp:ops | medium | Critical |
547,239,758 | pytorch | Reuse spawned subprocesses in RPC tests | ## 🚀 Enhancement
Investigating the possibility to reuse spawned subprocesses for `test_rpc_spawn.py` would be super helpful. The current test suite takes > 5 mins to complete, while the `test_rpc_fork` that we removed in https://github.com/pytorch/pytorch/pull/29827 completed much faster. I'm thus assuming that the reason for this time difference is due to spawning subprocesses in each test run.
It seems like it'd help developer productivity quite a bit if we were able to reuse these processes somehow across runs. Though we'd need to be careful to ensure that this does not introduce extra flakiness.
## Alternatives
Continue with per-test spawning.
cc @pietern @mrshenli @pritamdamania87 @zhaojuanmao @satgera @gqchen @aazzolini @rohan-varma @xush6528 @jjlilley | triaged,better-engineering,module: rpc | low | Minor |
547,250,603 | flutter | Widget testing is failed when tapping outside of the display area | <!-- Thank you for using Flutter!
If you are looking for support, please check out our documentation
or consider asking a question on Stack Overflow:
* https://flutter.dev/
* https://api.flutter.dev/
* https://stackoverflow.com/questions/tagged/flutter?sort=frequent
If you have found a bug or if our documentation doesn't have an answer
to what you're looking for, then fill our the template below. Please read
our guide to filing a bug first: https://flutter.dev/docs/resources/bug-reports
-->
## Steps to Reproduce
<!--
Please tell us exactly how to reproduce the problem you are running into.
Please attach a small application (ideally just one main.dart file) that
reproduces the problem. You could use https://gist.github.com/ for this.
If the problem is with your application's rendering, then please attach
a screenshot and explain what the problem is.
-->
This test will be failed without scroll the scroll view and displaying the button before `tester.tap`. `Finder` finds the button but `onPressed` event is not happened.
`Finder` finds widgets from tree of widgets and NOT depends on display state, but `tester.tap` depends on display state.
Currently, it is difficult to know why `onPressed` event is not happened, because `Finder` finds the button successfully.
``` dart
// main.dart
import 'package:flutter/material.dart';
void main() => runApp(MyApp());
class MyApp extends StatelessWidget {
@override
Widget build(BuildContext context) {
return MaterialApp(
home: TestPage(),
);
}
}
class TestPage extends StatefulWidget {
static const scrollViewKey = Key('scrollView');
static const buttonKey = Key('button');
@override
_TestPageState createState() => _TestPageState();
}
class _TestPageState extends State<TestPage> {
int count = 0;
@override
Widget build(BuildContext context) {
return Scaffold(
body: SingleChildScrollView(
key: TestPage.scrollViewKey,
child: Column(
children: <Widget>[
// Push out [Text] and [RaisedButton] with a long widget.
SizedBox(height: 800),
Text(count.toString()),
RaisedButton(
key: TestPage.buttonKey,
onPressed: () {
setState(() {
count = 1;
});
},
)
],
),
),
);
}
}
```
``` dart
// widget_test.dart
import 'package:flutter/material.dart';
import 'package:flutter_test/flutter_test.dart';
import 'package:scrollview_issue_example/main.dart';
void main() {
testWidgets('out of display test', (WidgetTester tester) async {
await tester.pumpWidget(MyApp());
expect(find.text('0'), findsOneWidget);
expect(find.text('1'), findsNothing);
// This test will be failed without dragging.
// await tester.drag(
// find.byKey(TestPage.scrollViewKey),
// const Offset(0.0, -300),
// );
// await tester.pump();
await tester.tap(find.byKey(TestPage.buttonKey));
await tester.pump();
expect(find.text('0'), findsNothing);
expect(find.text('1'), findsOneWidget);
});
}
```
<!-- Finally, paste the output of running `flutter doctor -v` here. -->
```
$ flutter doctor -v
[✓] Flutter (Channel stable, v1.12.13+hotfix.5, on Mac OS X 10.14.5 18F203, locale ja-JP)
• Flutter version 1.12.13+hotfix.5 at /Users/hiroki_matsue/work/flutter/flutter
• Framework revision 27321ebbad (4 weeks ago), 2019-12-10 18:15:01 -0800
• Engine revision 2994f7e1e6
• Dart version 2.7.0
...
• No issues found!
```
| a: tests,framework,f: gestures,has reproducible steps,P2,found in release: 3.3,found in release: 3.6,team-framework,triaged-framework | low | Critical |
547,256,240 | flutter | When popping a route, controls should reacquire focus without opening the keyboard or scrolling to make visible. | I have number of pages and all of them are kind of forms with some `TextFields`.
In each page, I am entering some data in `TextField` and then moving to the next page either by clicking next button directly (Key Board is Visible) or by clicking back button to dismiss keyboard then clicking next button. After the next page, When I trying to go to previous page by back button or `Navigator.pop` that `TextField` getting auto focused. This behaviour is really annoying lot.
I verified that in my older version app (Which uses older Flutter SDK) works fine.
**Sample code to reproduce:**
```
import 'package:flutter/material.dart';
void main() => runApp(MainPage());
class MainPage extends StatelessWidget {
@override
Widget build(BuildContext context) {
return MaterialApp(
home: FirstPage(),
);
}
}
class FirstPage extends StatelessWidget {
@override
Widget build(BuildContext context) {
double height = MediaQuery.of(context).size.height;
return Scaffold(
appBar: AppBar(
title: Text("First Page"),
),
body: SingleChildScrollView(
child: Column(
children: <Widget>[
TextField(),
Text("Use back button to dismiss keyboar"),
Text("Scroll down to see next button"),
SizedBox(
height: height + 100,
child: Center(child: Text("Some widgets")),
),
RaisedButton(
child: Text("Click Here"),
onPressed: () {
Navigator.push(
context,
MaterialPageRoute(
builder: (context) => SecondPage(),
),
);
},
)
],
),
),
);
}
}
class SecondPage extends StatelessWidget {
@override
Widget build(BuildContext context) {
return Scaffold(
appBar: AppBar(
title: Text("Second Page"),
),
body: Column(
children: <Widget>[
TextField(),
RaisedButton(
child: Text("Go Back"),
onPressed: () {
Navigator.pop(context);
},
)
],
),
);
}
}
```
**Flutter Doctor:**
```
[√] Flutter (Channel stable, v1.12.13+hotfix.5, on Microsoft Windows [Version 10.0.18362.535], locale en-IN)
• Flutter version 1.12.13+hotfix.5 at D:\Installed\FlutterSDK
• Framework revision 27321ebbad (4 weeks ago), 2019-12-10 18:15:01 -0800
• Engine revision 2994f7e1e6
• Dart version 2.7.0
[√] Android toolchain - develop for Android devices (Android SDK version 28.0.3)
• Android SDK at C:\Users\Flower\AppData\Local\Android\sdk
• Android NDK location not configured (optional; useful for native profiling support)
• Platform android-29, build-tools 28.0.3
• Java binary at: C:\Program Files\Android\Android Studio\jre\bin\java
• Java version OpenJDK Runtime Environment (build 1.8.0_202-release-1483-b03)
• All Android licenses accepted.
[√] Android Studio (version 3.5)
• Android Studio at C:\Program Files\Android\Android Studio
• Flutter plugin version 42.1.1
• Dart plugin version 191.8593
• Java version OpenJDK Runtime Environment (build 1.8.0_202-release-1483-b03)
[√] Connected device (1 available)
• KOB L09 • VAF4T18531000189 • android-arm64 • Android 7.0 (API 24)
• No issues found!
```
| a: text input,framework,f: routes,f: focus,has reproducible steps,P2,team-framework,triaged-framework,found in release: 3.19,found in release: 3.20 | low | Critical |
547,274,413 | flutter | Hero Animations with Popup Route |
## Steps to Reproduce
Created a custom page route to act / create a bottom sheet, doing this leads to Hero widget not working at all.
Created a dartPad example to showoff the problem ==> https://dartpad.dev/e79bd9adf6845857e0829ec6fe5a4a55
**Target Platform:** All
## Logs
<!-- Finally, paste the output of running `flutter doctor -v` here. -->
```
[√] Flutter (Channel master, v1.13.6-pre.16, on Microsoft Windows [Version 10.0.18363.535], locale en-US)
• Flutter version 1.13.6-pre.16 at C:\src\flutter
• Framework revision fcaf9c4070 (3 weeks ago), 2019-12-21 14:03:01 -0800
• Engine revision 33813929e3
• Dart version 2.8.0 (build 2.8.0-dev.0.0 886615d0f9)
[√] Android toolchain - develop for Android devices (Android SDK version 29.0.2)
• Android SDK at C:\Users\Dell\AppData\Local\Android\sdk
• Platform android-29, build-tools 29.0.2
• Java binary at: C:\Program Files\Android\Android Studio\jre\bin\java
• Java version OpenJDK Runtime Environment (build 1.8.0_202-release-1483-b03)
• All Android licenses accepted.
[√] Chrome - develop for the web
• Chrome at C:\Program Files (x86)\Google\Chrome\Application\chrome.exe
[√] Visual Studio - develop for Windows (Visual Studio Community 2019 16.4.0)
• Visual Studio at C:\Program Files (x86)\Microsoft Visual Studio\2019\Community
• Visual Studio Community 2019 version 16.4.29519.181
[√] Android Studio (version 3.5)
• Android Studio at C:\Program Files\Android\Android Studio
• Flutter plugin version 39.0.3
• Dart plugin version 191.8423
• Java version OpenJDK Runtime Environment (build 1.8.0_202-release-1483-b03)
[√] Connected device (4 available)
• Android SDK built for x86 • emulator-5554 • android-x86 • Android 10 (API 29) (emulator)
• Windows • Windows • windows-x64 • Microsoft Windows [Version 10.0.18363.535]
• Chrome • chrome • web-javascript • Google Chrome 78.0.3904.108
• Web Server • web-server • web-javascript • Flutter Tools
• No issues found!
PS C:\Users\Dell> flutter doctor -v
[√] Flutter (Channel master, v1.13.6-pre.16, on Microsoft Windows [Version 10.0.18363.535], locale en-US)
• Flutter version 1.13.6-pre.16 at C:\src\flutter
• Framework revision fcaf9c4070 (3 weeks ago), 2019-12-21 14:03:01 -0800
• Engine revision 33813929e3
• Dart version 2.8.0 (build 2.8.0-dev.0.0 886615d0f9)
```
| c: new feature,framework,a: animation,f: routes,P3,team-framework,triaged-framework | low | Major |
547,432,807 | rust | TcpStream bind before connect | We now have `TcpStream::connect` and `TcpStream::connect_timeout` to create a `TcpStream`.
But there is no option for us to do before calling `connect()`, in particular, `bind()` to the source address.
Proposal:
```rust
fn connect_with<A>(bind_addr: &SocketAddr, A: addr) -> io::Result<TcpStream>
where A: ToSocketAddrs
{
// ...
}
fn connect_with_timeout(bind_addr: &SocketAddr,
addr: &SocketAddr,
timeout: Duration)
-> io::Result<TcpStream>
{
// ...
}
``` | T-libs-api,C-feature-request | low | Major |
547,471,314 | godot | LineEdit last characters disappear with text align=right and using a Dynamic font with char_space>0 | **Godot version:**
Godot 3.1 release / 3.2-beta 5
**OS/device including version:**
Windows 10 / GTX 1060
**Issue description:**
I have used a DynamicFont with char_space=2 in a LineEdit.
If you set the alignment to left, center, or fill everything works fine, but after changing the aligment of the LineEdit text to right aligned the last characters of the text disappear.
After setting the char_space of the font back to 0 the full text of the LineEdit visible using right alignment.
**Steps to reproduce:**
Create a LineEdit control which has the text right aligned.
Add a DynamicFont to the LineEdit and change the char_space to any value bigger zero.
**Minimal reproduction project:**
[LineEditCharSpaceBug.zip](https://github.com/godotengine/godot/files/4040331/LineEditCharSpaceBug.zip)
| bug,confirmed,topic:gui | low | Critical |
547,481,140 | TypeScript | Narrow number literal types with comparison | ## Search Terms
constant number literals narrowing comparison
## Suggestion
Narrow number literal type not only with (non)equality, but also with less-than etc.
This request is only about throwing away non-matching literals, and not about full-blown `number` narrowing.
## Use Cases
This could be used to direct narrowing with tuple unions.
## Examples
With literals
```ts
type X = 0 | 1 | 2 | 3;
function f(): X {
return 1;
}
function g(x: 0 | 3) {
console.log(x);
}
let x: X = f();
g(x); // Error, 1 | 2 is unexpected
if (x > 2) {
g(x); // Same error, but should be safe
}
if (x != 1 && x != 2) {
g(x); // Ok
}
```
With tuples
```ts
type X = [] | [string, string];
function f(): X { return [];}
function g(x: string) {
console.log(x);
}
let x: X = f();
g(x[0]); // Error, undefined is unexpected
if (x.length > 0) {
g(x[0]); // Error, but should be ok
}
if (x.length == 2) {
g(x[0]); // Ok
}
```
## Checklist
My suggestion meets these guidelines:
* [x] This wouldn't be a breaking change in existing TypeScript/JavaScript code
* [x] This wouldn't change the runtime behavior of existing JavaScript code
* [x] This could be implemented without emitting different JS based on the types of the expressions
* [x] This isn't a runtime feature (e.g. library functionality, non-ECMAScript syntax with JavaScript output, etc.)
* [x] This feature would agree with the rest of [TypeScript's Design Goals](https://github.com/Microsoft/TypeScript/wiki/TypeScript-Design-Goals).
| Suggestion,In Discussion | low | Critical |
547,545,186 | vscode | Git - 'Revert selected ranges' should work on individual lines, not hunks | - VS Code version: Code 1.41.1 (26076a4de974ead31f97692a0d32f90d735645c0, 2019-12-18T15:04:31.999Z)
- OS version: Linux x64 4.15.0-72-generic (Ubuntu 16)
- Does this issue occur when all extensions are disabled?: Yes
`Revert selected ranges` reverts any hunks in which there is a cursor. This is annoying when one wants more fine-grained control. Especially seeing as it is already easy to revert by hunk through other features.
Note that this is contrary to `Stage selected ranges`, which correctly stages only the selected lines | bug,help wanted,git | low | Major |
547,569,026 | PowerToys | The ProjectTemplate doesn't show up when using the search box in VS | It's also missing the required tags:
```
<LanguageTag>C++</LanguageTag>
<PlatformTag>windows</PlatformTag>
<ProjectTypeTag>extension</ProjectTypeTag>
``` | Issue-Bug,Area-Project Template | low | Major |
547,609,834 | pytorch | Trying to disable cuda to run torch on OSX run it on CPU |
```
anaconda3/envs/meshcnn/lib/python3.6/site-packages/torch/cuda/__init__.py", line 95, in _check_driver
raise AssertionError("Torch not compiled with CUDA enabled")
AssertionError: Torch not compiled with CUDA enabled
```
Seems this is where I need to tell it to use the CPU but kinda stuck
in this file
lib/python3.6/site-packages/torch/cuda/__init__.py
```
def _check_driver():
if not hasattr(torch._C, '_cuda_isDriverSufficient'):
raise AssertionError("Torch not compiled with CUDA enabled")
if not torch._C._cuda_isDriverSufficient():
if torch._C._cuda_getDriverVersion() == 0:
# found no NVIDIA driver on the system
raise AssertionError("""
Found no NVIDIA driver on your system. Please check that you
have an NVIDIA GPU and installed a driver from
http://www.nvidia.com/Download/index.aspx""")
else:
# TODO: directly link to the alternative bin that needs install
raise AssertionError("""
The NVIDIA driver on your system is too old (found version {}).
Please update your GPU driver by downloading and installing a new
version from the URL: http://www.nvidia.com/Download/index.aspx
Alternatively, go to: https://pytorch.org to install
a PyTorch version that has been compiled with your version
of the CUDA driver.""".format(str(torch._C._cuda_getDriverVersion())))
```
Any guidance?
cc @ngimel | module: build,module: cuda,triaged | low | Critical |
547,614,218 | rust | Unhelpful highlighting in type inference error | MWE of a type inference error that came up in a real project:
https://play.rust-lang.org/?version=stable&mode=debug&edition=2018&gist=cd3198d1a30439a01ca2cfa3a78b1124
The user should fix this error by adding an explicit type for the {integer} literal `1` (i.e. 1u32 or 1u64), since these could do completely different things for the generic type `T`. The compiler catches this, but the highlighting seems to imply that it can't infer the type of the entire expression `!(number >> 1)`, when in reality the problem is with the `1`: no matter what type `1` takes, the full expression is always type `T`. This is pretty confusing | C-enhancement,A-diagnostics,T-compiler,A-inference,D-confusing | low | Critical |
547,622,043 | flutter | Support data: URIs in Image.network | ## Use case
For testing purposes, we produce data samples which contain data URIs for pre-loaded image URLs. This ensures that the tests are fast and hermetic.
We'd like to transparently pass these data URIs to Image.network as this means we don't have to modify the production code just for testing purposes (and try to find all the occurrences of Image.network is not a trivial task).
## Proposal
Detect data URIs in Image.network, decode them and transparently call Image.memory instead. | a: tests,c: new feature,framework,customer: dream (g3),a: images,P3,team-framework,triaged-framework | low | Major |
547,634,123 | rust | Type inference can't resolve type of output of `dyn Future` inside of `impl Fn` inside of Vec | This code ([Playground](https://play.rust-lang.org/?version=stable&mode=debug&edition=2018&gist=2fb9d4046178e9fd1d5f61c2ec80b034))
```rust
fn get_closure_vec(length: usize) -> Vec<impl Fn() -> Box<dyn std::future::Future<Output = usize>>> {
vec![|| Box::new(async { 5 }) as Box<_>; length]
}
```
gives error
```
error[E0271]: type mismatch resolving `<impl std::future::Future as std::future::Future>::Output == usize`
--> src/lib.rs:2:13
|
2 | vec![|| Box::new(async { 5 }) as Box<_>; width]
| ^^^^^^^^^^^^^^^^^^^^^^^^^^ expected usize, found i32
|
= note: required for the cast to the object type `dyn std::future::Future<Output = usize>`
```
and this code ([Playground](https://play.rust-lang.org/?version=stable&mode=debug&edition=2018&gist=990b46471a01fa442ea6106ca6853be8))
```rust
fn get_closure_vec(length: usize) -> Vec<impl Fn() -> Box<dyn std::future::Future<Output = usize>>> {
vec![|| Box::new(async { 5 }); length]
}
```
also gives error
```
error[E0271]: type mismatch resolving `<[closure@src/lib.rs:2:10: 2:39] as std::ops::FnOnce<()>>::Output == std::boxed::Box<(dyn std::future::Future<Output = usize> + 'static)>`
--> src/lib.rs:1:41
|
1 | fn get_closure_vec(width: usize) -> Vec<impl Fn() -> Box<dyn std::future::Future<Output = usize>>> {
| ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ expected opaque type, found trait std::future::Future
|
= note: expected type `std::boxed::Box<impl std::future::Future>`
found type `std::boxed::Box<(dyn std::future::Future<Output = usize> + 'static)>`
= note: the return type of a function must have a statically known size
```
while this one works fine ([Playground](https://play.rust-lang.org/?version=stable&mode=debug&edition=2018&gist=6815a4f55394544fd78a18b446099d89))
```rust
fn get_closure() -> impl Fn() -> Box<dyn std::future::Future<Output = usize>> {
|| Box::new(async { 5 })
}
``` | T-lang,T-compiler,A-inference,A-impl-trait,C-bug | low | Critical |
547,635,684 | rust | `repr(C)` is not respected for enums with uninhabited variants | Uninhabited `repr(C)` enums are size zero, _not_ the size of their discriminant. [For instance](https://play.rust-lang.org/?version=stable&mode=debug&edition=2018&gist=ab603e5a7bb2d0894ea22595bfdf599e):
```rust
enum Never {}
#[repr(C, u64)]
enum E {
_X(Never),
_Y(Never),
}
fn main (){
use core::mem::size_of;
let never_size = size_of::<Never>();
assert_eq!(0, never_size);
let disr_size = size_of::<u64>();
assert_eq!(8, disr_size);
let expected_enum_size = disr_size + never_size;
let actual_enum_size = size_of::<E>();
assert_eq!(expected_enum_size, actual_enum_size); // FAIL! 8 != 0
}
```
At best, this leads to nonsensical errors. For instance, [this][infinite-size]:
```rust
enum Never {}
#[repr(C, u64)]
enum E {
_X(Never),
_Y(E),
}
```
[produces this error message](infinite-size):
```
error[E0072]: recursive type `E` has infinite size
--> src/lib.rs:4:1
|
4 | enum E {
| ^^^^^^ recursive type has infinite size
5 | _X(Never),
6 | _Y(E),
| - recursive without indirection
|
= help: insert indirection (e.g., a `Box`, `Rc`, or `&`) at some point to make `E` representable
```
[infinite-size]: https://play.rust-lang.org/?version=stable&mode=debug&edition=2018&gist=ef95b35c6f9951ee82edfda0ab6e8f6a
At worst, it violates the specification of [RFC2195: _Really Tagged Unions_](https://rust-lang.github.io/rfcs/2195-really-tagged-unions.html), which dictates [that `E` should be layout-compatible to the `union` of its variants](https://play.rust-lang.org/?version=stable&mode=debug&edition=2018&gist=86bf6b539516cc82280ddedb15730c45):
```rust
use core::mem::size_of;
#[derive(Copy, Clone)]
enum Never {}
// This:
#[repr(C, u64)]
enum E {
_X(Never),
_Y(Never),
}
// ...should have the same layout as `ERepr`:
#[derive(Copy, Clone)]
#[repr(C)]
union ERepr {
A: EX,
B: EY,
}
type ETag = u64;
#[derive(Copy, Clone)] struct EX(ETag, Never);
#[derive(Copy, Clone)] struct EY(ETag, Never);
// ...but it doesn't:
fn main() {
assert_eq!(size_of::<E>(), size_of::<ERepr>()); // FAIL: 0 != 8
}
```
| T-compiler,C-bug,T-opsem,A-repr | low | Critical |
547,642,648 | svelte | Portal in examples section | Can we add this Portal repl in the examples section?
https://github.com/sveltejs/svelte/issues/3088#issuecomment-505785516
```svelte
<script>
// src/components/Portal.svelte
import { onMount, onDestroy } from 'svelte'
let ref
let portal
onMount(() => {
portal = document.createElement('div')
portal.className = 'portal'
document.body.appendChild(portal)
portal.appendChild(ref)
})
onDestroy(() => {
document.body.removeChild(portal)
})
</script>
<div class="portal-clone">
<div bind:this={ref}>
<slot></slot>
</div>
</div>
<style>
.portal-clone { display: none; }
</style>
```
Copyright @ThomasJuster. | temp-stale,documentation | low | Major |
547,642,653 | godot | `ResourceSaver.save` error in tool script: `get_multiple_md5: Condition ' !f ' is true.` | **Godot version:**
Godot Engine v3.1.2.stable.official
**OS/device including version:**
Mac OS 10.14.6
**Issue description:**
Usage of `ResourceSaver.save(name, texture)` produce sometime the following error :
`
ERROR: get_multiple_md5: Condition ' !f ' is true. Continuing..:
At: core/os/file_access.cpp:622.
`
Which crash Godot (freeze)
**Steps to reproduce:**
1 - Unzip test project
2 - Start from command line
3 - Freeze of Godot, error in the terminal
Look into addons/codeandweb.texturepacker/texturepacker_import_spritesheet.gd line 115, it's where the method call is.
You can also see the that file here : [file on github](https://github.com/CodeAndWeb/texturepacker-godot-plugin/blob/master/addons/codeandweb.texturepacker/texturepacker_import_spritesheet.gd#L115)
**Minimal reproduction project:**
[importBug.zip](https://github.com/godotengine/godot/files/4041854/importBug.zip)
To be noted that I've already spent 3 hours with the plugin author to try to debug the issue, but our conclusion is that it's a bug inside Godot and we are a bit clueless on how to move forward to fix that :D
It doesn't seems to be an issue with the plugin itself, but if so please let met know where.
Thanks ! | bug,topic:core,confirmed,crash | low | Critical |
547,668,770 | terminal | Investigate independent input source for Swap Chain Panel | While reading some background documentation on how this whole co* stuff worked, I noticed that `SwapChainPanel` apparently has a provision to let background threads potentially mess with the chain instead of having to jump to the UI thread to do so:
https://docs.microsoft.com/en-us/uwp/api/windows.ui.xaml.controls.swapchainpanel.createcoreindependentinputsource#Windows_UI_Xaml_Controls_SwapChainPanel_CreateCoreIndependentInputSource_Windows_UI_Core_CoreInputDeviceTypes_
It's probably not easily/instantly applicable here. But I want to leave the note because I thought of it and @DHowett-MSFT and @zadjii-msft might have an opinion on whether that would be a decent follow-on task to generate from this comment.
_Originally posted by @miniksa in https://github.com/microsoft/terminal/pull/4051/threads/MDIzOlB1bGxSZXF1ZXN0UmV2aWV3VGhyZWFkMjIzNTY3ODg3OnYy/unresolve_
This represents investigating whether using this is a good idea or not. | Help Wanted,Area-Rendering,Product-Terminal,Issue-Task | low | Minor |
547,674,602 | opencv | CV_Assert does not properly use global namespace lookups | If a user does something like this:
```c++
namespace X
{
namespace cv
{
auto
my_function(::cv::InputArray img) -> ::cv::Mat_<uchar>
{
CV_Assert(img.channels() == 3);
}
}
}
```
,
the code will fail to compile with failed lookup of `cv::Error`.
```bash
error: no member named 'Error' in namespace 'X::cv'; did you mean '::cv::Error'?
CV_Assert(img.channels() == 3);
```
This is because of how the macro expands out by using `cv` without global scope resolution via `::cv`. | bug,priority: low,category: build/install | low | Critical |
547,686,231 | flutter | PaginatedDataTable footer likely won't handle text zoom correctly | As pointed out by @Hixie [here](https://github.com/flutter/flutter/pull/47555/files/b3deb9139d0df48a7df442afcf03c94dedb3b487#diff-243d50d685c9399394efa4693f2ac703R488), the footer for a `PaginatedDataTable` likely won't handle text zoom correctly. | framework,f: material design,team-design,triaged-design | low | Minor |
547,693,576 | go | x/sys/windows: Some types should be available for non-Windows GOOS targets | <!--
Please answer these questions before submitting your issue. Thanks!
For questions please use one of our forums: https://github.com/golang/go/wiki/Questions
-->
### What version of Go are you using (`go version`)?
<pre>
$ go version
go version go1.13.5 darwin/amd64
</pre>
### Does this issue reproduce with the latest release?
Yes.
### What operating system and processor architecture are you using (`go env`)?
<details><summary><code>go env</code> Output</summary><br><pre>
$ go env
GO111MODULE=""
GOARCH="amd64"
GOBIN=""
GOCACHE="/Users/[snip]/Library/Caches/go-build"
GOENV="/Users/[snip]/Library/Application Support/go/env"
GOEXE=""
GOFLAGS=""
GOHOSTARCH="amd64"
GOHOSTOS="darwin"
GONOPROXY="[snip]"
GONOSUMDB="[snip]"
GOOS="darwin"
GOPATH="/Users/[snip]/go"
GOPRIVATE="[snip]"
GOPROXY="https://proxy.golang.org,direct"
GOROOT="/usr/local/Cellar/go/1.13.5/libexec"
GOSUMDB="sum.golang.org"
GOTMPDIR=""
GOTOOLDIR="/usr/local/Cellar/go/1.13.5/libexec/pkg/tool/darwin_amd64"
GCCGO="gccgo"
AR="ar"
CC="clang"
CXX="clang++"
CGO_ENABLED="1"
GOMOD=""
CGO_CFLAGS="-g -O2"
CGO_CPPFLAGS=""
CGO_CXXFLAGS="-g -O2"
CGO_FFLAGS="-g -O2"
CGO_LDFLAGS="-g -O2"
PKG_CONFIG="pkg-config"
GOGCCFLAGS="-fPIC -m64 -pthread -fno-caret-diagnostics -Qunused-arguments -fmessage-length=0 -fdebug-prefix-map=/var/folders/84/_6l41bt970l9fmsrwc_p1gv00000gn/T/go-build550252566=/tmp/go-build -gno-record-gcc-switches -fno-common"
</pre></details>
### What did you do?
<!--
If possible, provide a recipe for reproducing the error.
A complete runnable program is good.
A link on play.golang.org is best.
-->
Attempt to use `golang.org/x/sys/windows.GUID` on a non-Windows OS. If you're not running Windows, this will do it:
```golang
package main
import "github.com/Microsoft/go-winio/pkg/guid"
func main() {
guid.ParseString("not a guid")
}
```
Assuming that file is saved as `main.go`, run: `go run main.go`.
(`github.com/Microsoft/go-winio/pkg/guid.GUID` [is defined as](https://github.com/Microsoft/go-winio/blob/v0.4.14/pkg/guid/guid.go#L47) `golang.org/x/sys/windows.GUID`)
### What did you expect to see?
The command would run and exit successfully, without an error or other message.
### What did you see instead?
This compile error:
`[snip]/github.com/Microsoft/go-winio/pkg/guid/guid.go:16:2: build constraints exclude all Go files in [snip]/golang.org/x/sys/windows`
### Additional information
I was attempting to use this to create something to interface with ActiveDirectory (or LDAP), but won't be running on Windows. There are common values in ActiveDirectory that are GUIDs, so it would be very helpful to be able to use this type. Especially in combination with [Microsoft's package](https://pkg.go.dev/github.com/Microsoft/go-winio/pkg/guid?tab=doc) that uses that type. | OS-Windows,NeedsDecision,compiler/runtime | medium | Critical |
547,708,367 | pytorch | gradients inside gradient checkpoint | ## 🚀 Feature
allowing the use of `torch.autorgrad.grad` and `loss.backward` inside `torch.utils.checkpoint.checkpoint`
## Motivation
Enclosing the entire training steps inside a checkpoint could improve significantly the memory cost for the gradient computation in large scale hyperparameter optimization (HO) and meta-learning algorithms (ML) based on unrolled differentiation. Here you have to backpropagate over the entire training procedure and doing that naively with pytorch has a really high cost in memory because it records all the activations and all the weights for all the training steps. Using checkpointing you could just store the weights for all the training steps and recompute the activation and the weight update operation (which include the gradient of the loss) during the backward pass.
## Alternatives
The alternative to use checkpointing is to implement a custom backward pass for these use cases exploiting the particular structure of the problem to save memory. I personally think that the requested feature provides a much simpler solution.
cc @ezyang @SsnL @albanD @zou3519 @gqchen | module: checkpoint,module: autograd,triaged,enhancement | low | Minor |
547,722,501 | godot | Using `discard` in a shader has different outcome depending on where it is placed | Godot 3.2 beta5
Windows 10 64 bits
OpenGL ES 3.0 Renderer: GeForce GTX 1060 6GB/PCIe/SSE2
When `discard` is used to fade opaque objects, the result is either correct or noisy depending on where the `discard` is placed. I thought it did not matter because the result is not getting rendered, but apparently it still had an effect.
At first I wanted to place it at the beginning, thinking it was best place to cancel the rest of operations early. But it produces noise (enlarge picture to see details, even on the sky):
But if I move it at the end, all is fine:
I tried this with a `SpatialMaterial` using the dithering distance fade, it also reproduced. It's unclear to me wether this would be a Godot bug, driver bug, or some GPU specificity.
Project:
[DiscardFading.zip](https://github.com/godotengine/godot/files/4042471/DiscardFading.zip)
| bug,topic:rendering,confirmed | low | Critical |
547,739,985 | kubernetes | DeletionTimestamp set after grace period is respected does not include grace period time | Dear Kubernetes-Team:
I ran an experiment with an application that does not exit after Kubelet sends an inital SIGTERM signal.
The specified grace-period is respected and the pod is deleted after the grace period finishes.
The deletionTimestamp on the inital delete event seems to include the grace-period, but the deletionTimestamp that is finally updated by the APIServer does not account for the grace-period that was respected. This DeletionTimestamp is also seen when the DeleteFunc is triggered on separate podInformers we have running on our clusters.
I have attached some test results detailing this below; This behavior seems like a bug. The final DeletionTimestamp that is updated by the APIServer should indeed include any grace-period that was respected. (Ideally it should be when the application exits, and not just at the time the user puts in the delete request)
My experiment is as follows:
1) Create a busybox pod with a termination grace period (for example 5 mins)
2) Call delete pod
3) APIServer gets a delete event initiated by the user (k8s-admin) with a deletion timestamp by adding the 5 mins grace period
4) After the grace period ends api server gets another delete event from kubelet (system:node:fargate-ip-19)
5) This delete event from kubelet has grace period of 0 and the DeletionTimestamp set to when user deleted the pod (instead of that time + the respected grace period)
The podspec `busybox.yaml` that is applied:
```
apiVersion: v1
kind: Pod
metadata:
name: busybox
namespace: default
spec:
containers:
- image: busybox
command:
- sleep
- "3600"
imagePullPolicy: IfNotPresent
name: busybox
restartPolicy: Always
terminationGracePeriodSeconds: 300
```
I executed:
`kubectl get pods -oyaml -w` after performing a `kubectl delete -f busybox.yaml` which gives us:
```
apiVersion: v1
kind: Pod
metadata:
annotations:
kubernetes.io/psp: eks.privileged
creationTimestamp: "2020-01-09T20:49:15Z"
deletionGracePeriodSeconds: 300
deletionTimestamp: "2020-01-09T20:56:25Z"
labels:
eks.amazonaws.com/fargate-profile: test-fp-pritrame
name: busybox
namespace: default
resourceVersion: "8941315"
selfLink: /api/v1/namespaces/default/pods/busybox
uid: 7f91ad24-3321-11ea-80df-0622abc9f058
spec:
containers:
- command:
- sleep
- "3600"
image: busybox
imagePullPolicy: IfNotPresent
name: busybox
resources: {}
terminationMessagePath: /dev/termination-log
terminationMessagePolicy: File
volumeMounts:
- mountPath: /var/run/secrets/kubernetes.io/serviceaccount
name: default-token-jpf24
readOnly: true
dnsPolicy: ClusterFirst
enableServiceLinks: true
nodeName: fargate-ip-10-0-45-109.us-west-2.compute.internal
priority: 2000001000
priorityClassName: system-node-critical
restartPolicy: Always
schedulerName: fargate-scheduler
securityContext: {}
serviceAccount: default
serviceAccountName: default
terminationGracePeriodSeconds: 300
tolerations:
- effect: NoExecute
key: node.kubernetes.io/not-ready
operator: Exists
tolerationSeconds: 300
- effect: NoExecute
key: node.kubernetes.io/unreachable
operator: Exists
tolerationSeconds: 300
volumes:
- name: default-token-jpf24
secret:
defaultMode: 420
secretName: default-token-jpf24
status:
conditions:
- lastProbeTime: null
lastTransitionTime: "2020-01-09T20:49:55Z"
status: "True"
type: Initialized
- lastProbeTime: null
lastTransitionTime: "2020-01-09T20:50:00Z"
status: "True"
type: Ready
- lastProbeTime: null
lastTransitionTime: "2020-01-09T20:50:00Z"
status: "True"
type: ContainersReady
- lastProbeTime: null
lastTransitionTime: "2020-01-09T20:49:55Z"
status: "True"
type: PodScheduled
containerStatuses:
- containerID: containerd://3d4a21fd691b182a54d9a4bc07ee4a94af7f35c5a7a54e26c127fa9673b57f48
image: docker.io/library/busybox:latest
imageID: docker.io/library/busybox@sha256:6915be4043561d64e0ab0f8f098dc2ac48e077fe23f488ac24b665166898115a
lastState: {}
name: busybox
ready: true
restartCount: 0
state:
running:
startedAt: "2020-01-09T20:49:59Z"
hostIP: 10.0.45.109
phase: Running
podIP: 10.0.45.109
qosClass: BestEffort
startTime: "2020-01-09T20:49:55Z"
---
apiVersion: v1
kind: Pod
metadata:
annotations:
kubernetes.io/psp: eks.privileged
creationTimestamp: "2020-01-09T20:49:15Z"
deletionGracePeriodSeconds: 300
deletionTimestamp: "2020-01-09T20:56:25Z"
labels:
eks.amazonaws.com/fargate-profile: test-fp-pritrame
name: busybox
namespace: default
resourceVersion: "8941926"
selfLink: /api/v1/namespaces/default/pods/busybox
uid: 7f91ad24-3321-11ea-80df-0622abc9f058
spec:
containers:
- command:
- sleep
- "3600"
image: busybox
imagePullPolicy: IfNotPresent
name: busybox
resources: {}
terminationMessagePath: /dev/termination-log
terminationMessagePolicy: File
volumeMounts:
- mountPath: /var/run/secrets/kubernetes.io/serviceaccount
name: default-token-jpf24
readOnly: true
dnsPolicy: ClusterFirst
enableServiceLinks: true
nodeName: fargate-ip-10-0-45-109.us-west-2.compute.internal
priority: 2000001000
priorityClassName: system-node-critical
restartPolicy: Always
schedulerName: fargate-scheduler
securityContext: {}
serviceAccount: default
serviceAccountName: default
terminationGracePeriodSeconds: 300
tolerations:
- effect: NoExecute
key: node.kubernetes.io/not-ready
operator: Exists
tolerationSeconds: 300
- effect: NoExecute
key: node.kubernetes.io/unreachable
operator: Exists
tolerationSeconds: 300
volumes:
- name: default-token-jpf24
secret:
defaultMode: 420
secretName: default-token-jpf24
status:
conditions:
- lastProbeTime: null
lastTransitionTime: "2020-01-09T20:49:55Z"
status: "True"
type: Initialized
- lastProbeTime: null
lastTransitionTime: "2020-01-09T20:56:26Z"
message: 'containers with unready status: [busybox]'
reason: ContainersNotReady
status: "False"
type: Ready
- lastProbeTime: null
lastTransitionTime: "2020-01-09T20:56:26Z"
message: 'containers with unready status: [busybox]'
reason: ContainersNotReady
status: "False"
type: ContainersReady
- lastProbeTime: null
lastTransitionTime: "2020-01-09T20:49:55Z"
status: "True"
type: PodScheduled
containerStatuses:
- image: busybox
imageID: ""
lastState: {}
name: busybox
ready: false
restartCount: 0
state:
waiting:
reason: ContainerCreating
hostIP: 10.0.45.109
phase: Pending
podIP: 10.0.45.109
qosClass: BestEffort
startTime: "2020-01-09T20:49:55Z"
---
apiVersion: v1
kind: Pod
metadata:
annotations:
kubernetes.io/psp: eks.privileged
creationTimestamp: "2020-01-09T20:49:15Z"
deletionGracePeriodSeconds: 0
deletionTimestamp: "2020-01-09T20:51:25Z"
labels:
eks.amazonaws.com/fargate-profile: test-fp-pritrame
name: busybox
namespace: default
resourceVersion: "8941941"
selfLink: /api/v1/namespaces/default/pods/busybox
uid: 7f91ad24-3321-11ea-80df-0622abc9f058
spec:
containers:
- command:
- sleep
- "3600"
image: busybox
imagePullPolicy: IfNotPresent
name: busybox
resources: {}
terminationMessagePath: /dev/termination-log
terminationMessagePolicy: File
volumeMounts:
- mountPath: /var/run/secrets/kubernetes.io/serviceaccount
name: default-token-jpf24
readOnly: true
dnsPolicy: ClusterFirst
enableServiceLinks: true
nodeName: fargate-ip-10-0-45-109.us-west-2.compute.internal
priority: 2000001000
priorityClassName: system-node-critical
restartPolicy: Always
schedulerName: fargate-scheduler
securityContext: {}
serviceAccount: default
serviceAccountName: default
terminationGracePeriodSeconds: 300
tolerations:
- effect: NoExecute
key: node.kubernetes.io/not-ready
operator: Exists
tolerationSeconds: 300
- effect: NoExecute
key: node.kubernetes.io/unreachable
operator: Exists
tolerationSeconds: 300
volumes:
- name: default-token-jpf24
secret:
defaultMode: 420
secretName: default-token-jpf24
status:
conditions:
- lastProbeTime: null
lastTransitionTime: "2020-01-09T20:49:55Z"
status: "True"
type: Initialized
- lastProbeTime: null
lastTransitionTime: "2020-01-09T20:56:26Z"
message: 'containers with unready status: [busybox]'
reason: ContainersNotReady
status: "False"
type: Ready
- lastProbeTime: null
lastTransitionTime: "2020-01-09T20:56:26Z"
message: 'containers with unready status: [busybox]'
reason: ContainersNotReady
status: "False"
type: ContainersReady
- lastProbeTime: null
lastTransitionTime: "2020-01-09T20:49:55Z"
status: "True"
type: PodScheduled
containerStatuses:
- image: busybox
imageID: ""
lastState: {}
name: busybox
ready: false
restartCount: 0
state:
waiting:
reason: ContainerCreating
hostIP: 10.0.45.109
phase: Pending
podIP: 10.0.45.109
qosClass: BestEffort
startTime: "2020-01-09T20:49:55Z"
---
apiVersion: v1
kind: Pod
metadata:
annotations:
kubernetes.io/psp: eks.privileged
creationTimestamp: "2020-01-09T20:49:15Z"
deletionGracePeriodSeconds: 0
deletionTimestamp: "2020-01-09T20:51:25Z"
labels:
eks.amazonaws.com/fargate-profile: test-fp-pritrame
name: busybox
namespace: default
resourceVersion: "8941942"
selfLink: /api/v1/namespaces/default/pods/busybox
uid: 7f91ad24-3321-11ea-80df-0622abc9f058
spec:
containers:
- command:
- sleep
- "3600"
image: busybox
imagePullPolicy: IfNotPresent
name: busybox
resources: {}
terminationMessagePath: /dev/termination-log
terminationMessagePolicy: File
volumeMounts:
- mountPath: /var/run/secrets/kubernetes.io/serviceaccount
name: default-token-jpf24
readOnly: true
dnsPolicy: ClusterFirst
enableServiceLinks: true
nodeName: fargate-ip-10-0-45-109.us-west-2.compute.internal
priority: 2000001000
priorityClassName: system-node-critical
restartPolicy: Always
schedulerName: fargate-scheduler
securityContext: {}
serviceAccount: default
serviceAccountName: default
terminationGracePeriodSeconds: 300
tolerations:
- effect: NoExecute
key: node.kubernetes.io/not-ready
operator: Exists
tolerationSeconds: 300
- effect: NoExecute
key: node.kubernetes.io/unreachable
operator: Exists
tolerationSeconds: 300
volumes:
- name: default-token-jpf24
secret:
defaultMode: 420
secretName: default-token-jpf24
status:
conditions:
- lastProbeTime: null
lastTransitionTime: "2020-01-09T20:49:55Z"
status: "True"
type: Initialized
- lastProbeTime: null
lastTransitionTime: "2020-01-09T20:56:26Z"
message: 'containers with unready status: [busybox]'
reason: ContainersNotReady
status: "False"
type: Ready
- lastProbeTime: null
lastTransitionTime: "2020-01-09T20:56:26Z"
message: 'containers with unready status: [busybox]'
reason: ContainersNotReady
status: "False"
type: ContainersReady
- lastProbeTime: null
lastTransitionTime: "2020-01-09T20:49:55Z"
status: "True"
type: PodScheduled
containerStatuses:
- image: busybox
imageID: ""
lastState: {}
name: busybox
ready: false
restartCount: 0
state:
waiting:
reason: ContainerCreating
hostIP: 10.0.45.109
phase: Pending
podIP: 10.0.45.109
qosClass: BestEffort
startTime: "2020-01-09T20:49:55Z"
```
We see above that the final updated DeletionTimestamp is set to: `2020-01-09T20:51:25Z` which is the time I put my initial delete request in, and which does not account for the grace-period, even though the grace period was respected.
We see the initial DeletionTimestamp set to: `2020-01-09T20:56:25Z` with a 300 second `terminationGracePeriod` set.
Here is a log line from on of our pod Informers with the timestamp showing when the DeleteFunc was triggered for this pod, which shows that the grace period of 5 minutes was respected:
```I0109 20:56:34.083187 1 eventhandlers.go:147] Received delete event for pod: default/busybox```
| kind/bug,priority/backlog,sig/node,lifecycle/frozen,triage/accepted | medium | Critical |
547,746,512 | flutter | Consider changing the Color.lerp implementation to more naturally handle transparency | Simply run with dart pad and press the FAB button. I'm not show is AnimatedContainer problem or SKIA colour rendering problem.
``` dart
import 'package:flutter/material.dart';
void main() => runApp(MyApp());
class MyApp extends StatelessWidget {
@override
Widget build(BuildContext context) {
return MaterialApp(
title: 'Transparent Issue',
debugShowCheckedModeBanner: false,
theme: ThemeData(
primarySwatch: Colors.blue,
),
home: MyHomePage(title: 'Colors.transparent issue'),
);
}
}
class MyHomePage extends StatefulWidget {
MyHomePage({Key key, this.title}) : super(key: key);
final String title;
@override
_MyHomePageState createState() => _MyHomePageState();
}
class _MyHomePageState extends State<MyHomePage> {
bool isColorSwap = true;
void _incrementCounter() {
setState(() {
isColorSwap = !isColorSwap;
});
}
@override
Widget build(BuildContext context) {
return Scaffold(
backgroundColor: Colors.white,
appBar: AppBar(
title: Text(widget.title),
),
body: Column(
crossAxisAlignment: CrossAxisAlignment.stretch,
children: <Widget>[
Expanded(
child: AnimatedContainer(
child: Center(child: Text('fade between white color')),
width: double.infinity,
duration: Duration(milliseconds: 500),
color: isColorSwap ? Colors.lightBlue[100] : Colors.white)),
Expanded(
child: AnimatedContainer(
child:
Center(child: Text('fade between transparent color')),
width: double.infinity,
duration: Duration(milliseconds: 500),
color: isColorSwap
? Colors.lightBlue[100]
: Colors.transparent))
]),
floatingActionButton: FloatingActionButton(
onPressed: _incrementCounter,
tooltip: 'ClickMe',
child: Icon(Icons.add),
),
);
}
}
``` | engine,a: animation,c: API break,has reproducible steps,P2,found in release: 3.3,found in release: 3.6,team-engine,triaged-engine | low | Critical |
547,774,564 | terminal | Add more `CommandlineTest` tests for `wt.exe` commandlines | Test excessively long text. Test empty text. Test text with ridiculous Unicode points in it like RTL/LTR/ZWJ/BOM. Test with invalid Unicode characters like the 0x80 0xc0 version of null.
_Originally posted by @miniksa in https://github.com/microsoft/terminal/pull/4023_ | Help Wanted,Product-Terminal,Issue-Task,Area-CodeHealth | low | Minor |
547,774,908 | terminal | Make a fuzzer for the `wt.exe` commandlines | We should probably make a fuzzer for this.
_Originally posted by @miniksa in https://github.com/microsoft/terminal/pull/4023_ | Help Wanted,Product-Terminal,Issue-Task,Area-CodeHealth | low | Minor |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.