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 |
|---|---|---|---|---|---|---|
491,301,648 | go | crypto/tls: Dial error message does not give much context | ### What version of Go are you using (`go version`)?
<pre>
go version go1.11.13 linux/amd64
</pre>
### What operating system and processor architecture are you using (`go env`)?
<details><summary><code>go env</code> Output</summary><br><pre>
GOARCH="amd64"
GOBIN=""
GOCACHE="/home/mbirc/.cache/go-build"
GOEXE=""
GOFLAGS=""
GOHOSTARCH="amd64"
GOHOSTOS="linux"
GOOS="linux"
GOPATH="/home/mbirc/go"
GOPROXY=""
GORACE=""
GOROOT="/usr/lib64/go/1.11"
GOTMPDIR=""
GOTOOLDIR="/usr/lib64/go/1.11/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-build241684230=/tmp/go-build -gno-record-gcc-switches"
</pre></details>
### What did you do?
<details><summary>tls-client.go code</summary><br><pre>
package main
import (
"crypto/tls"
"os"
"log"
)
func main() {
args:= os.Args
config := tls.Config{}
conn, err := tls.Dial("tcp", args[1] + ":" + args[2], &config)
if err != nil {
log.Fatalf("client: dial: %s", err)
}
defer conn.Close()
log.Println("client: connected to: ", conn.RemoteAddr())
}
</pre></details>
This does not work for the successful case but in error case it prints:
<pre>
client: dial: remote error: tls: handshake failure
</pre>
### What did you expect to see?
Message that allows diagnosing the failure. Why has it failed to connect? Other clients connect fine.
### What did you see instead?
Unexplained failure. | NeedsInvestigation | low | Critical |
491,350,758 | pytorch | Incorrect lable read with ImageInput Op of Caffe2 | ## 🐛 Bug
Incorrect label data on reading mnist lmdb with image_input operator. I see the same behavior with both C++ and python interface. mnist db read works as expected if I use TensorProtosDBInput operator.
## To Reproduce
Steps to reproduce the behavior:
1. Please run the script below to reproduce. You will need to add full path to data_folder variable for mnist db
<!-- If you have a code sample, error messages, stack traces, please provide it here as well -->
import os, sys
import shutil
from caffe2.python import core, model_helper, net_drawer, workspace, visualize, brew
from caffe2.python.helpers.tools import image_input
# This section preps your image and test set in a lmdb database
def DownloadResource(url, path):
'''Downloads resources from s3 by url and unzips them to the provided path'''
import requests, zipfile, StringIO
print("Downloading... {} to {}".format(url, path))
try:
r = requests.get(url, stream=True)
except Exception as ex:
print("Failed to download dataset. Please download it manually from {}".format(url))
print("Unzip it and place the two database folders here: {}".format(path))
raise ex
z = zipfile.ZipFile(StringIO.StringIO(r.content))
z.extractall(path)
print("Completed download and extraction.")
# If you would like to see some really detailed initializations,
# you can change --caffe2_log_level=0 to --caffe2_log_level=-1
core.GlobalInit(['caffe2', '--caffe2_log_level=0'])
print("Necessities imported!")
data_folder = "./"
root_folder = os.path.join("./",'tutorial_mnist')
if not os.path.exists(data_folder):
os.makedirs(data_folder)
print("Your data folder was not found!! This was generated: {}".format(data_folder))
# Look for existing database: lmdb
if os.path.exists(os.path.join(data_folder,"mnist-train-nchw-lmdb")):
print("lmdb train db found!")
else:
db_missing = True
if os.path.exists(os.path.join(data_folder,"mnist-test-nchw-lmdb")):
print("lmdb test db found!")
else:
db_missing = True
if os.path.exists(root_folder):
print("Looks like you ran this before, so we need to cleanup those old files...")
shutil.rmtree(root_folder)
os.makedirs(root_folder)
workspace.ResetWorkspace(root_folder)
print("training data folder:" + data_folder)
print("workspace root folder:" + root_folder)
def AddInput(model, batch_size, db, db_type,useTensorProto=True):
# load the data
if useTensorProto:
data_uint8, label = model.TensorProtosDBInput([], ["data_uint8", "label"], batch_size=batch_size,db=db, db_type=db_type)
else:
data_uint8, label = image_input(model,[],["data_uint8", "label"],batch_size=batch_size,
db=db, db_type=db_type, is_test=1,scale=28,color=0,crop=28,use_caffe_datum=0,label_type=0,
num_decode_threads=1)
# cast the data to float
data = model.Cast(data_uint8, "data", to=core.DataType.FLOAT)
# scale data from [0,255] down to [0,1]
data = model.Scale(data, data, scale=float(1./256))
# don't need the gradient for the backward pass
data = model.StopGradient(data, data)
return data, label
def main(argv):
arg_scope = {"order": "NCHW"}
## Load images using TensorProtoDBInput
train_model = model_helper.ModelHelper(name="mnist_train-TensorProto", arg_scope=arg_scope)
data, label = AddInput(
train_model, batch_size=64,
db=os.path.join(data_folder,"mnist-train-nchw-lmdb"),
db_type='lmdb',useTensorProto=True)
print(label)
workspace.RunNetOnce(train_model.param_init_net)
# creating the network
workspace.CreateNet(train_model.net, overwrite=True)
# Run Net
workspace.RunNet(train_model.net)
labelData=workspace.FetchBlob('label')
print(labelData)
## Load images using ImageInput
workspace.ResetWorkspace(root_folder)
train_model = model_helper.ModelHelper(name="mnist_train-image_input", arg_scope=arg_scope)
data, label = AddInput(
train_model, batch_size=64,
db=os.path.join(data_folder,"mnist-train-nchw-lmdb"),
db_type='lmdb',useTensorProto=False)
print(label)
workspace.RunNetOnce(train_model.param_init_net)
# creating the network
workspace.CreateNet(train_model.net, overwrite=True)
# Run Net
workspace.RunNet(train_model.net)
labelData=workspace.FetchBlob('label')
print(labelData)
sys.exit(0)
if __name__ == '__main__':
main(sys.argv[1:])
## Expected behavior
Output on ruining the script. Bold label data is from Tensorprotodb op where as the italic label data is from image_input
E0909 14:56:37.372220 215429 init_intrinsics_check.cc:43] CPU feature avx is present on your machine, but the Caffe2 binary is not compiled with it. It means you may not get the full speed of your CPU.
E0909 14:56:37.372248 215429 init_intrinsics_check.cc:43] CPU feature avx2 is present on your machine, but the Caffe2 binary is not compiled with it. It means you may not get the full speed of your CPU.
E0909 14:56:37.372252 215429 init_intrinsics_check.cc:43] CPU feature fma is present on your machine, but the Caffe2 binary is not compiled with it. It means you may not get the full speed of your CPU.
Necessities imported!
Looks like you ran this before, so we need to cleanup those old files...
training data folder:./
workspace root folder:./tutorial_mnist
**label
[5 0 4 1 9 2 1 3 1 4 3 5 3 6 1 7 2 8 6 9 4 0 9 1 1 2 4 3 2 7 3 8 6 9 0 5 6
0 7 6 1 8 7 9 3 9 8 5 9 3 3 0 7 4 9 8 0 9 4 1 4 4 6 0]**
label
E0909 14:56:37.384630 215429 image_input_op.h:250] You are using an old ImageInputOp format that creates a local db reader. Consider moving to the new style that takes in a DBReader blob instead.
I0909 14:56:37.385756 215429 image_input_op.h:325] Creating an image input op with the following setting:
I0909 14:56:37.385769 215429 image_input_op.h:326] Using 4 CPU threads;
I0909 14:56:37.385772 215429 image_input_op.h:330] Outputting in batches of 64 images;
I0909 14:56:37.385776 215429 image_input_op.h:331] Treating input image as grayscale image;
I0909 14:56:37.385779 215429 image_input_op.h:345] Scaling image to 28 without warping;
I0909 14:56:37.385782 215429 image_input_op.h:359] Central cropping image to 28 without random mirroring;
I0909 14:56:37.385785 215429 image_input_op.h:362] Label Type: 0
I0909 14:56:37.385789 215429 image_input_op.h:363] Num Labels: 0
I0909 14:56:37.385792 215429 image_input_op.h:371] Default [Channel 0] Subtract mean 0 and divide by std 1.
I0909 14:56:37.385809 215429 image_input_op.h:371] Default [Channel 1] Subtract mean 0 and divide by std 1.
I0909 14:56:37.385813 215429 image_input_op.h:371] Default [Channel 2] Subtract mean 0 and divide by std 1.
I0909 14:56:37.385821 215429 image_input_op.h:376] Outputting images as unknown.
_[1953394531 543977330 543516788 1867525728 1869052268 1881170030
1701867378 1701409906 168442483 538976288 1867524133 1866692972
1667591276 1852795252 168457001 1970562386 175337074 757935405
170732845 1867525728 1866692972 1667591276 1852795252 538970720
541138976 1867525728 1866692972 1667591276 1852795252 1868767328
1767994478 1735289198 1701344288 1869377568 1684370548 1819242528
1852794745 168439411 543515987 1869835329 757935370 757935405
1768294957 1650420844 1702327397 544763493 1766203450 1646292076
1702327397 1948282469 1931505527 544437349 2015389295 1818326573
779314549 1867385354 175334772 757935405 774769197 1869503264
544433524 1952671091 544108393 1970365810]_
## Environment
Collecting environment information...
PyTorch version: 1.1.0a0+863818e
Is debug build: No
CUDA used to build PyTorch: 10.0.130
OS: Red Hat Enterprise Linux Server release 7.6 (Maipo)
GCC version: (GCC) 4.9.3
CMake version: version 2.8.12.2
Python version: 3.6
Is CUDA available: Yes
CUDA runtime version: 10.0.130
GPU models and configuration:
GPU 0: GeForce RTX 2080 Ti
GPU 1: GeForce RTX 2080 Ti
GPU 2: GeForce RTX 2080 Ti
GPU 3: GeForce RTX 2080 Ti
Nvidia driver version: 410.48
cuDNN version: Probably one of the following:
/appl/gpu-local.rhel.7.6/cuda-10.0/lib64/libcudnn.so.7
/usr/lib64/libcudnn.so.7.5.0
Versions of relevant libraries:
[pip3] numpy==1.16.0
[pip3] torch==1.1.0a0+863818e
[pip3] torchvision==0.2.1
[conda] Could not collect
## Additional context
<!-- Add any other context about the problem here. -->
| caffe2,triaged | low | Critical |
491,352,626 | flutter | Platform Views are massively expensive and the docs and instrumentation don't highlight this enough. | Platform views are massively expensive in terms on memory usage, loss of concurrency, battery usage, additional render target switches, etc.. Based on feedback seen from customers (for example, in this [internal doc](https://docs.google.com/document/d/16Vel2Nwy97Bc65yIiTpF-zevgArLqK8ISdgIQtV_igw/edit?ts=5d76b651#heading=h.r2i7gn60zi3s)), it is observed that customers may not be adequately aware of the negative performance implications of the addition of a platform views in their applications. While these impacts were known upfront, they are only documented at the lowest levels of the API at [`SceneBuilder::addPlatformView`](https://api.flutter.dev/flutter/dart-ui/SceneBuilder/addPlatformView.html). Even the documentation at that level is easy to miss. Also, customers rarely use that API directly (instead platform views are wrapped in higher level widgets) and may miss the moment they step off a steep performance cliff.
While the implementation of platform views continues to get better, the savings are minor compared to the baseline impact of the addition of a single platform view to the hierarchy. The well informed developer has the opportunity to make their applications significantly more performant and delightful for their users.
A deeper discussion of the performance impact of the feature should include a discussion of the following issues:
* The additional memory use.
* Discuss how the extra backing stores have be created.
* Discuss how large each backing store can get (screen size times buffer count which could be as high as 3 plus some additional fixed overhead).
* Discuss how the over is not constant and is function of the number of interleaving levels in the resultant view hierarchy.
* The negative implications to concurrency (subject to platform specific restrictions).
* How a more restrictive thread configuration needs to be chosen (no GPU thread).
* How the frame pipeline depth needs to be reduced.
* The cost of switching thread configurations (barrier blocks and such).
* De-optimizations in the renderer due to fewer retained layers from frame to frame.
* De-optimization in the driver due to additional render target switches.
* Discuss how expensive a render target switch is relative to something the user can observe (say a backdrop filter).
* The degradation in battery life.
* Discuss how Flutter has to perform more composition (that would otherwise have been avoided if there was no interleaving level).
* Discuss how the system compositor has to do more composition.
It is recommended that these discussions are surfaces at multiple levels in the documentation. Including the following landing pages specifically identified:
* The [PlatformViewLayer](https://api.flutter.dev/flutter/rendering/PlatformViewLayer-class.html) class reference and each memory of the same (for example the separate page for the constructor which comes up before everything else in search results).
* The specific link for the [platform view layer](https://api.flutter.dev/flutter/rendering/rendering-library.html) in the rendering library.
* It is recommended that discussions about performance be highlighted somehow (maybe a red highlight to draw attention so it is hard to miss).
* Highlight these platform view related discussions in the [Performance best practices](https://flutter.dev/docs/testing/best-practices) section. The omission of platform views from this section is alarming and we should address it ASAP. Especially since this feature has a more severe performance implication than everything currently on that page.
* Highlight how to find and audit the use of platform views in the application in the section on [performance profiling](https://flutter.dev/docs/testing/ui-performance). Again, this section is not accurate today w.r.t the discussion on threading configuration when a platform view is in use.
* In each of the sections on the main docs sites, a discussion of the performant alternatives to the use of platform views ought to be added.
I suspect these discussions will need to be quickly and constantly updated as the some of the performance impacts are ameliorated. Platform views are really powerful and sometimes the only way to achieve some key piece of functionality in the application. But a more informed approach to their use can significantly improve application performance along multiple dimensions. | framework,engine,d: api docs,a: platform-views,P3,a: plugins,team-engine,triaged-engine | low | Major |
491,359,160 | pytorch | Python/C++ API Parity: torch.nn modules and functional | Currently, PyTorch C++ API is missing many `torch::nn` layers that are available in the Python API. As part of the Python/C++ API parity work, we would like to add the following `torch::nn` modules and utilities in C++ API:
## Containers
- [ ] Module (**TODO**: some APIs are missing in C++, e.g. `register_forward_hook` / `register_forward_pre_hook`)
- [x] Sequential (@ShahriarSS)
- [x] ModuleList (@ShahriarSS)
- [x] ModuleDict (@ejguan https://github.com/pytorch/pytorch/pull/47707)
- [x] ParameterList (@yyn19951228 https://github.com/pytorch/pytorch/pull/41259)
- [x] ParameterDict (@yyn19951228 https://github.com/pytorch/pytorch/pull/40654)
## Convolution layers
- [x] Conv1d (@yf225 https://github.com/pytorch/pytorch/pull/28917)
- [x] Conv2d (@yf225 https://github.com/pytorch/pytorch/pull/28917)
- [x] Conv3d (@yf225 https://github.com/pytorch/pytorch/pull/28917)
- [x] ConvTranspose1d (@nuka137 https://github.com/pytorch/pytorch/pull/29721)
- [x] ConvTranspose2d (@nuka137 https://github.com/pytorch/pytorch/pull/29721)
- [x] ConvTranspose3d (@nuka137 https://github.com/pytorch/pytorch/pull/29721)
- [x] Unfold (@jon-tow https://github.com/pytorch/pytorch/pull/27809)
- [x] Fold (@ShahriarSS https://github.com/pytorch/pytorch/pull/24160)
## Pooling layers
- [x] MaxPool1d (@ShahriarSS https://github.com/pytorch/pytorch/pull/24860)
- [x] MaxPool2d (@ShahriarSS https://github.com/pytorch/pytorch/pull/24860)
- [x] MaxPool3d (@ShahriarSS https://github.com/pytorch/pytorch/pull/24860)
- [x] MaxUnpool1d (@pbelevich https://github.com/pytorch/pytorch/pull/26896)
- [x] MaxUnpool2d (@pbelevich https://github.com/pytorch/pytorch/pull/26915)
- [x] MaxUnpool3d (@pbelevich https://github.com/pytorch/pytorch/pull/27027)
- [x] AvgPool1d (@ShahriarSS https://github.com/pytorch/pytorch/pull/25800)
- [x] AvgPool2d (@ShahriarSS https://github.com/pytorch/pytorch/pull/25800)
- [x] AvgPool3d (@ShahriarSS https://github.com/pytorch/pytorch/pull/25800)
- [x] FractionalMaxPool2d (@yf225 https://github.com/pytorch/pytorch/pull/29933)
- [x] FractionalMaxPool3d (@yf225 https://github.com/pytorch/pytorch/pull/29933)
- [x] LPPool1d (@nuka137 https://github.com/pytorch/pytorch/pull/27800)
- [x] LPPool2d (@nuka137 https://github.com/pytorch/pytorch/pull/28492)
- [x] AdaptiveMaxPool1d (@pbelevich https://github.com/pytorch/pytorch/pull/26755)
- [x] AdaptiveMaxPool2d (@pbelevich https://github.com/pytorch/pytorch/pull/26772)
- [x] AdaptiveMaxPool3d (@pbelevich https://github.com/pytorch/pytorch/pull/26775)
- [x] AdaptiveAvgPool1d (@pbelevich https://github.com/pytorch/pytorch/pull/26808)
- [x] AdaptiveAvgPool2d (@pbelevich https://github.com/pytorch/pytorch/pull/26818)
- [x] AdaptiveAvgPool3d (@pbelevich https://github.com/pytorch/pytorch/pull/26819)
## Padding layers
- [x] ReflectionPad1d (@yf225 https://github.com/pytorch/pytorch/pull/28538)
- [x] ReflectionPad2d (@yf225 https://github.com/pytorch/pytorch/pull/28538)
- [x] ReplicationPad1d (@yf225 https://github.com/pytorch/pytorch/pull/28539)
- [x] ReplicationPad2d (@yf225 https://github.com/pytorch/pytorch/pull/28539)
- [x] ReplicationPad3d (@yf225 https://github.com/pytorch/pytorch/pull/28539)
- [x] ZeroPad2d (@yf225 https://github.com/pytorch/pytorch/pull/28540)
- [x] ConstantPad1d (@yf225 https://github.com/pytorch/pytorch/pull/28541)
- [x] ConstantPad2d (@yf225 https://github.com/pytorch/pytorch/pull/28541)
- [x] ConstantPad3d (@yf225 https://github.com/pytorch/pytorch/pull/28541)
## Non-linear activations (weighted sum, nonlinearity)
- [x] ELU (@pbelevich https://github.com/pytorch/pytorch/pull/27028)
- [x] Hardshrink (@pbelevich https://github.com/pytorch/pytorch/pull/27035)
- [x] Hardtanh (@pbelevich https://github.com/pytorch/pytorch/pull/27038)
- [x] LeakyReLU (@pbelevich https://github.com/pytorch/pytorch/pull/27059)
- [x] LogSigmoid (@pbelevich https://github.com/pytorch/pytorch/pull/27060)
- [x] MultiheadAttention (@pbelevich https://github.com/pytorch/pytorch/pull/27309)
- [x] PReLU (@pbelevich https://github.com/pytorch/pytorch/pull/27429)
- [x] ReLU (@pbelevich https://github.com/pytorch/pytorch/pull/27435)
- [x] ReLU6 (@pbelevich https://github.com/pytorch/pytorch/pull/27436)
- [x] RReLU (@pbelevich https://github.com/pytorch/pytorch/pull/27437)
- [x] SELU (@jon-tow https://github.com/pytorch/pytorch/pull/27434)
- [x] CELU (@pbelevich https://github.com/pytorch/pytorch/pull/27487)
- [x] GELU (@BIT-silence https://github.com/pytorch/pytorch/pull/28944)
- [x] Sigmoid (@pbelevich https://github.com/pytorch/pytorch/pull/27488)
- [x] Softplus (@pbelevich https://github.com/pytorch/pytorch/pull/27489)
- [x] Softshrink (@pbelevich https://github.com/pytorch/pytorch/pull/27534)
- [x] Softsign (@pbelevich https://github.com/pytorch/pytorch/pull/27535)
- [x] Tanh (@pbelevich https://github.com/pytorch/pytorch/pull/27536)
- [x] Tanhshrink (@pbelevich https://github.com/pytorch/pytorch/pull/27537)
- [x] Threshold (@pbelevich https://github.com/pytorch/pytorch/pull/27538)
- [x] GLU (@yf225 https://github.com/pytorch/pytorch/pull/29922)
- [x] SiLU (@heitorschueroff https://github.com/pytorch/pytorch/pull/41034)
## Non-linear activations (other)
- [x] Softmin (@nuka137 https://github.com/pytorch/pytorch/pull/27459)
- [x] Softmax (@nuka137 https://github.com/pytorch/pytorch/pull/27446)
- [x] Softmax2d (@nuka137 https://github.com/pytorch/pytorch/pull/27509)
- [x] LogSoftmax (@nuka137 https://github.com/pytorch/pytorch/pull/27462)
- [x] AdaptiveLogSoftmaxWithLoss (@mansoorcheema https://github.com/pytorch/pytorch/pull/29076)
## Normalization layers
- [x] BatchNorm1d (@nuka137 https://github.com/pytorch/pytorch/pull/28176)
- [x] BatchNorm2d (@nuka137 https://github.com/pytorch/pytorch/pull/28936)
- [x] BatchNorm3d (@nuka137 https://github.com/pytorch/pytorch/pull/28936)
- [x] GroupNorm (@yf225 https://github.com/pytorch/pytorch/pull/29920)
- [ ] SyncBatchNorm (Needs distributed data parallel support)
- [x] InstanceNorm1d (@divyanshsinghvi https://github.com/pytorch/pytorch/pull/28790)
- [x] InstanceNorm2d (@divyanshsinghvi https://github.com/pytorch/pytorch/pull/28790)
- [x] InstanceNorm3d (@divyanshsinghvi https://github.com/pytorch/pytorch/pull/28790)
- [x] LayerNorm (@anjali411 https://github.com/pytorch/pytorch/pull/28032)
- [x] LocalResponseNorm (@mansoorcheema https://github.com/pytorch/pytorch/pull/28759)
- [x] CrossMapLRN2d (@lsrock1 https://github.com/pytorch/pytorch/pull/29039)
## Recurrent layers
- [x] RNN (@yf225 https://github.com/pytorch/pytorch/pull/34322)
- [x] LSTM (@yf225 https://github.com/pytorch/pytorch/pull/34322)
- [x] GRU (@yf225 https://github.com/pytorch/pytorch/pull/34322)
- [x] RNNCell (@yf225 https://github.com/pytorch/pytorch/pull/34400)
- [x] LSTMCell (@yf225 https://github.com/pytorch/pytorch/pull/34400)
- [x] GRUCell (@yf225 https://github.com/pytorch/pytorch/pull/34400)
## Transformer layers
- [x] Transformer (@glaringlee https://github.com/pytorch/pytorch/pull/44333)
- [x] TransformerEncoder (@glaringlee https://github.com/pytorch/pytorch/pull/43187)
- [x] TransformerDecoder (@VinodSKumar https://github.com/pytorch/pytorch/pull/42886)
- [x] TransformerEncoderLayer (@glaringlee https://github.com/pytorch/pytorch/pull/42633)
- [x] TransformerDecoderLayer (@VinodSKumar https://github.com/pytorch/pytorch/pull/42717)
## Linear layers
- [x] Identity (@jon-tow https://github.com/pytorch/pytorch/pull/26713)
- [x] Linear (@pbelevich https://github.com/pytorch/pytorch/pull/27382)
- [x] Bilinear (@MJ10 https://github.com/pytorch/pytorch/pull/26082)
- [x] Flatten (@mrsalehi https://github.com/pytorch/pytorch/pull/28072)
- [x] Unflatten (@heitorschueroff https://github.com/pytorch/pytorch/pull/42613)
## Dropout layers
- [x] Dropout (@pbelevich https://github.com/pytorch/pytorch/pull/29761)
- [x] Dropout2d (@pbelevich https://github.com/pytorch/pytorch/pull/29761)
- [x] Dropout3d (@pbelevich https://github.com/pytorch/pytorch/pull/29761)
- [x] AlphaDropout (@Suyash458 https://github.com/pytorch/pytorch/pull/28424)
- [x] FeatureAlphaDropout (@Suyash458 https://github.com/pytorch/pytorch/pull/28424)
## Sparse layers
- [x] Embedding (@anjali411 https://github.com/pytorch/pytorch/pull/26358)
- [x] EmbeddingBag (@anjali411 https://github.com/pytorch/pytorch/pull/26358)
## Distance functions
- [x] CosineSimilarity (@jon-tow https://github.com/pytorch/pytorch/pull/26424)
- [x] PairwiseDistance (@jon-tow https://github.com/pytorch/pytorch/pull/26424)
## Loss functions
- [x] L1Loss (@ShahriarSS https://github.com/pytorch/pytorch/pull/25902)
- [x] MSELoss (@ShahriarSS https://github.com/pytorch/pytorch/pull/27156)
- [x] CrossEntropyLoss (@PyExtreme https://github.com/pytorch/pytorch/pull/29812)
- [x] CTCLoss (@pbelevich https://github.com/pytorch/pytorch/pull/28654)
- [x] NLLLoss (@PyExtreme https://github.com/pytorch/pytorch/pull/29812)
- [x] PoissonNLLLoss (@pbelevich https://github.com/pytorch/pytorch/pull/28755)
- [x] KLDivLoss (@ShahriarSS https://github.com/pytorch/pytorch/pull/27156)
- [x] BCELoss (@ShahriarSS https://github.com/pytorch/pytorch/pull/27156)
- [x] BCEWithLogitsLoss (@pbelevich https://github.com/pytorch/pytorch/pull/28783)
- [x] MarginRankingLoss (@pbelevich https://github.com/pytorch/pytorch/pull/29000)
- [x] HingeEmbeddingLoss (@jon-tow https://github.com/pytorch/pytorch/pull/27101)
- [x] MultiLabelMarginLoss (@CarMiranda https://github.com/pytorch/pytorch/pull/27659)
- [x] SmoothL1Loss (@CarMiranda https://github.com/pytorch/pytorch/pull/27661)
- [x] SoftMarginLoss (@CarMiranda https://github.com/pytorch/pytorch/pull/27660)
- [x] MultiLabelSoftMarginLoss (@CarMiranda https://github.com/pytorch/pytorch/pull/27669)
- [x] CosineEmbeddingLoss (@jon-tow https://github.com/pytorch/pytorch/pull/27345)
- [x] MultiMarginLoss (please see details in https://github.com/pytorch/pytorch/issues/27198) (@CarMiranda https://github.com/pytorch/pytorch/pull/27424)
- [x] TripletMarginLoss (please see details in https://github.com/pytorch/pytorch/issues/27197) (@PyExtreme https://github.com/pytorch/pytorch/pull/27713)
## Vision layers
- [x] PixelShuffle (@CarMiranda https://github.com/pytorch/pytorch/pull/28140)
- [x] Upsample (along with `torch::nn::functional::interpolate`) (@jon-tow https://github.com/pytorch/pytorch/pull/28413)
## Utilities
- [x] clip_grad_norm_ (@rohan-varma https://github.com/pytorch/pytorch/pull/26140)
- [x] clip_grad_value_ (@jokerkeny https://github.com/pytorch/pytorch/pull/28736)
- [x] parameters_to_vector (@lsrock1 https://github.com/pytorch/pytorch/pull/29267)
- [x] vector_to_parameters (@lsrock1 https://github.com/pytorch/pytorch/pull/29267)
- [ ] weight_norm (@bernsm3, has dependency on `Module._forward_pre_hooks`)
- [ ] remove_weight_norm (@bernsm3, has dependency on `Module._forward_pre_hooks`)
- [ ] spectral_norm (has dependency on `Module._forward_pre_hooks`)
- [ ] remove_spectral_norm (has dependency on `Module._forward_pre_hooks`)
- [x] PackedSequence (@yf225 https://github.com/pytorch/pytorch/pull/33652)
- [x] pack_padded_sequence (@yf225 https://github.com/pytorch/pytorch/pull/33652)
- [x] pad_packed_sequence (@yf225 https://github.com/pytorch/pytorch/pull/33652)
- [x] pad_sequence (@Suyash458 https://github.com/pytorch/pytorch/pull/32387)
- [x] pack_sequence (@yf225 https://github.com/pytorch/pytorch/pull/33652)
## torch.nn.functional
- [x] normalize (please see details in https://github.com/pytorch/pytorch/issues/27048) (@gtamba https://github.com/pytorch/pytorch/pull/27280)
- [x] gumbel_softmax (please see details in https://github.com/pytorch/pytorch/issues/27078) (@Naresh1318 https://github.com/pytorch/pytorch/pull/28121)
- [x] one_hot (please see details in https://github.com/pytorch/pytorch/issues/27081) (@bzinodev https://github.com/pytorch/pytorch/pull/27177)
- [x] pdist (please see details in https://github.com/pytorch/pytorch/issues/27082) (@jon-tow https://github.com/pytorch/pytorch/pull/27122)
- [x] grid_sample (@lsrock1 https://github.com/pytorch/pytorch/pull/28354)
- [x] affine_grid (please see details in https://github.com/pytorch/pytorch/issues/27196) (@jon-tow https://github.com/pytorch/pytorch/pull/27263)
- [x] gelu (@anjali411 https://github.com/pytorch/pytorch/pull/28433)
## Implementation Notes:
- For `torch.nn` modules, Python and C++ implementation must only differ in language-specific syntax, and their data member fields, control flow and logic must be **exactly** the same.
- You might see that some of the `torch.nn` modules call the corresponding `torch.nn.functional` functions. When implementing the C++ version of those modules, we should also add the corresponding `torch::nn::functional` functions, in order to preserve the call structure.
- When you are adding a new C++ `torch::nn` module:
- The new `torch::nn` module must subclass from `public Cloneable<ModuleName>`. For example, `class TORCH_API LinearImpl : public Cloneable<LinearImpl>` (in `torch/csrc/api/include/torch/nn/modules/linear.h`).
- How do we decide which file to put the new C++ `torch::nn` module in? Answer: We should look at where the Python version of module is located and try to mirror that file structure. For example, the Python version of `torch.nn.Linear` lives in `torch/nn/modules/linear.py`, so the C++ version `torch::nn::Linear` should live in `torch/csrc/api/include/torch/nn/modules/linear.h`.
- If you add a new module header, you must also add the include statement for the new header in `torch/csrc/api/include/torch/nn/modules.h`.
- If you add a new module `.cpp` file, you must also add it into `caffe2/CMakeLists.txt` and `tools/build_variables.py`. (Please search for other modules in those files to see how this should be done.)
- You must add tests for the new module in `test/cpp/api/modules.cpp`. In particular, make sure the module's `pretty_print` is tested and it outputs the same value as the Python version.
- if the module is templatized (e.g. `MaxPoolImpl`), you must add explicit template instantiation in the module’s `.cpp` file (e.g. search for `template class` in `torch/csrc/api/src/nn/modules/pooling.cpp` to see how this is done).
- When you are adding a new C++ `torch::nn::functional` function:
- How do we decide which file to put the new C++ `torch::nn::functional` function? Answer: We should look at where the corresponding `torch::nn` module is located. For example, `torch::nn::PairwiseDistance` lives in `torch/csrc/api/include/torch/nn/modules/distance.h` (which, following the rule above, is determined by `torch/nn/modules/distance.py`), so `torch::nn::functional::pairwise_distance` should live in `torch/csrc/api/include/torch/nn/functional/distance.h`.
- You must add tests for the new functional in `test/cpp/api/functional.cpp`.
- If you add a new header file for the functional, you must add include statement for the header file in `torch/csrc/api/include/torch/nn/functional.h`.
- The name of the functional must match the Python version.
- When you are adding a new module/functional options:
- If the functional options only has optional arguments, the functional's function signature should have `options = {}`, to allow users to call the functional without passing options. And we must add a test for this.
- If the functional / module options only has one non-optional argument, we need the implicit constructor for the options that only takes the non-optional argument (i.e. `/* implicit */ Options(value_type value);`), so that we are able to do `functional(input, options_arg_value)` / `auto m = Module(options_arg_value)` instead of `functional(input, Options(options_arg_value))` / `auto m = Module(Options(options_arg_value))`. And we must add a test for this. (See `DropoutOptions` in `torch/csrc/api/include/torch/nn/options/dropout.h` as example.)
- for optional Tensor argument, we should use `torch::Tensor` (with `Tensor()` as default “null" value), and don’t use `c10::optional<Tensor>`.
- If the options is templatized (e.g. `MaxPoolOptions` in `torch/csrc/api/include/torch/nn/options/pooling.h`), you must add explicit template instantiation in the options’ `.cpp` file (search for `template struct` in `torch/csrc/api/src/nn/options/pooling.cpp` to see how this is done).
- Do not use `torch::IntArrayRef`, use `std::vector<int64_t>` instead.
## How do I run tests?
- Follow https://github.com/pytorch/pytorch#from-source to build your branch of PyTorch from source.
- To test `test/cpp/api/modules.cpp`, run `./build/bin/test_api --gtest_filter=ModulesTest* --gtest_stack_trace_depth=10 --gmock_verbose=info`.
- To test `test/cpp/api/functional.cpp`, run `./build/bin/test_api --gtest_filter=FunctionalTest* --gtest_stack_trace_depth=10 --gmock_verbose=info`.
Please ask for @yf225's review when you open a PR to add a `torch::nn` module from this list.
Tracking post on PyTorch forum: https://discuss.pytorch.org/t/55650
cc @yf225
| module: cpp,module: nn,good first issue,triaged | high | Critical |
491,379,532 | godot | MIDI Input in iOS is not implemented | **Godot version:**
3.1
**OS/device including version:**
iPhone 6, iOS 11.2.5
**Issue description:**
This code should produce a result when input is received from a MIDI-interface. It has been confirmed to work on desktop. For some reason it does not work on iOS.
```
OS.open_midi_inputs()
OS.get_connected_midi_inputs()
set_process_unhandled_input(true)
func _unhandled_input(event : InputEvent):
if (event is InputEventMIDI):
if (event.get_channel() == 1):
print ("MIDI Action Received")
```
**Steps to reproduce:**
Use above code on iOS device with MIDI-interface attached. (Or use the example below that should react to input from a MIDI device sending on channel 1.)
**Minimal reproduction project:**
[MIDI-Example.zip](https://github.com/godotengine/godot/files/3593079/MIDI-Example.zip)
| bug,platform:ios,confirmed,topic:audio,topic:input | low | Minor |
491,388,440 | pytorch | Forward/backward hooks for C++ torch::nn modules | ## 🚀 Feature
As part of the Python/C++ API parity work, we would like to provide forward/backward hooks for C++ `torch::nn` modules:
(left: Python API, right: C++ API)
- `torch.nn.Module.register_forward_pre_hook` -> `torch::nn::Module::register_forward_pre_hook`
- `torch.nn.Module.register_forward_hook` -> `torch::nn::Module::register_forward_hook`
- `torch.nn.Module.register_backward_hook` -> `torch::nn::Module::register_backward_hook`
## Possible Steps
- Implement `torch::utils::hooks::RemovableHandle` in C++ API, which mirrors `torch.utils.hooks.RemovableHandle` in Python API.
- Implement `register_forward_pre_hook`, `register_forward_hook` and `register_backward_hook` methods for `torch::nn::Module`.
- We might need to add `_backward_hooks`, `_forward_pre_hooks` and `_forward_hooks` to `torch::nn::Module` (torch/csrc/api/include/torch/nn/module.h).
- Add appropriate C++ tests for the new API, similar to Python tests.
cc @ezyang @SsnL @albanD @zou3519 @yf225 | module: cpp,module: autograd,triaged | medium | Critical |
491,416,441 | vscode | removeSecondaryCursors should leave the cursor added last | `removeSecondaryCursors` (i.e. clicking "Escape" after selecting multiple cursors), should leave the cursor which was added last. This allows fast searching for other uses in the file (`ctrl-d ctrl-d esc` as opposed to `ctrl-d ctrl-c ctrl-f ctrl-v enter esc`). | feature-request,editor-multicursor,editor-commands | low | Minor |
491,420,415 | pytorch | load_state_dict on CPU first | ## 🚀 Feature
When calling load_state_dict on model and optimizer, we should load to CPU first then copy to GPU
## Motivation
load_state_dict currently has a call to deepcopy which doubles the RAM for those weights. Thus when a model is already pushing the limit of RAM it cannot be loaded back in memory.
## Pitch
If we load on CPU first, the deepcopy can happen on CPU memory before going to GPU
## Alternatives
Building smaller models
@ezyang | module: serialization,triaged | low | Minor |
491,467,154 | TypeScript | Incorrect error message on export assignment in `.d.ts` file | **TypeScript Version:** 3.5.3
**Search Terms:**
An export assignment cannot be used in a namespace.
**Code**
This is pretty niche, but I think it has general implications...
The current `/lib/typescript.d.ts` file is modelled as essentially a CommonJS file where at the end of the file, it has `export = ts;` which exports the whole namespace. The actual `/lib/typescript.js` though will place itself in the global scope as `ts` if it isn't loaded as a CommonJS module. In order to express that `ts` is available as a global variable, it requires code like this to be placed in a `.d.ts` file:
```ts
import * as _ts from "typescript";
declare global {
namespace ts {
export = _ts;
}
}
```
Other methods of expressing this, like `declare var ts: typeof _ts;` only make the function shape of `ts` available, while the interfaces and type aliases and other type information is not accessible.
**Expected behavior:**
No errors.
**Actual behavior:**
The error `TS1063: An export assignment cannot be used in a namespace` occurs, but in VSCode, if the file is named `globals.d.ts` this error does not occur. 🤷♂ In all cases the code above does work like expected, making both the functional and type exports of the namespace available for code to be checked against.
| Suggestion,In Discussion | low | Critical |
491,508,537 | pytorch | Python hang after using torch.exp() | ## 🐛 Bug
Using function torch.exp() makes python hang. I test some other functions like torch.sum(), torch.abs(), these functions can return right results.
<!-- A clear and concise description of what the bug is. -->
## To Reproduce
Steps to reproduce the behavior:
```
torch.exp(torch.Tensor([0, math.log(2)]))
```
Python just hang, python cannot be closed by using CTRL+C under Ubuntu 16.04.
<!-- If you have a code sample, error messages, stack traces, please provide it here as well -->
## Expected behavior
Out[2]: Tensor([1, 2])
<!-- A clear and concise description of what you expected to happen. -->
## Environment
Collecting environment information...
PyTorch version: 1.2.0
Is debug build: No
CUDA used to build PyTorch: 10.0.130
OS: Ubuntu 16.04.5 LTS
GCC version: (Ubuntu 5.3.1-14ubuntu2) 5.3.1 20160413
CMake version: version 3.5.1
Python version: 3.6
Is CUDA available: Yes
CUDA runtime version: 9.0.176
GPU models and configuration:
GPU 0: GeForce GTX 1080 Ti
GPU 1: GeForce GTX 1080 Ti
Nvidia driver version: 410.93
cuDNN version: Could not collect
Versions of relevant libraries:
[pip] numpy==1.16.2
[pip] numpydoc==0.9.1
[pip] torch==1.2.0
[pip] torch-cluster==1.4.2
[pip] torch-geometric==1.2.1
[pip] torch-scatter==1.2.0
[pip] torch-sparse==0.4.0
[pip] torch-spline-conv==1.1.0
[pip] torchsummaryX==1.3.0
[pip] torchvision==0.4.0a0+6b959ee
[conda] blas 1.0 mkl
[conda] cuda90 1.0 h6433d27_0 pytorch
[conda] mkl 2018.0.3 1
[conda] mkl-service 1.1.2 py36h90e4bf4_5
[conda] torch 1.2.0 pypi_0 pypi
[conda] torch-cluster 1.4.2 pypi_0 pypi
[conda] torch-geometric 1.2.1 pypi_0 pypi
[conda] torch-scatter 1.2.0 pypi_0 pypi
[conda] torch-sparse 0.4.0 pypi_0 pypi
[conda] torch-spline-conv 1.1.0 pypi_0 pypi
[conda] torchsummaryx 1.3.0 pypi_0 pypi
[conda] torchvision 0.2.2.post2 pypi_0 pypi
| needs reproduction,triaged,module: deadlock | low | Critical |
491,540,148 | flutter | Don't rely on global `pod` installation, use bundled version | ## Use case
In addition to stock Cococapods we have some plugins required in our Podfile.
Cocoapods itself and the plugins are version controlled using `bundler`, and when running `bundle exec pod install` in the `ios/` directory everything works fine.
But when executing `flutter run` it tried to use the global `pod` binary, which doesn't pick up the locally installed plugin we want to use.
## Proposal
I am not super familiar with Cocoapods, so there might be a better solution to this.
What I would propose is to remove the hardcoded `pod install` in places like this: https://github.com/flutter/flutter/blob/b7c714e84c1e085e068ec1b8b2b18d4cbada76d0/packages/flutter_tools/lib/src/macos/cocoapods.dart#L280
…and instead do some discovery to find out whether pod is locally available via `bundle exec` and in that case use that instead.
If there is any other workaround to make `pod install` aware of the local plugins that would also be fine.
I think in general it would be much better though not even to require the global `pod` installation, and do everything through bundled versions. (In our team we constantly have noise due to different people upgrading Cocoapods at different times and hence modifying the version string in the `Podfile.lock` file) | platform-ios,tool,customer: crowd,c: proposal,P3,team-ios,triaged-ios | medium | Critical |
491,558,407 | TypeScript | Automated Migration for Breaking Changes to the Type System | Automated Migration for Breaking Changes to the Type System
## Search Terms
codemod, migration, upgrade
## Suggestion
semi-automated migration when there are breaking changes in the type system
One way of implementing this would be via codemods (could use the TS Compiler API) that do trivial update tasks as described in https://github.com/microsoft/TypeScript/issues/33272
## Use Cases
- Would make it easier for users to upgrade to the latest TypeScript
- Would make it easier for the TS team to upgrade TypeScript
- Might enable TS to evolve faster, as the pain from these "breaking" changes will be smaller
The changes described in https://github.com/microsoft/TypeScript/issues/33272, for example, were mostly trivial, and ideally would not need much human intervention.
Type-only changes are especially amenable to this kind of treatment, because no one's app will misbehave at runtime if there is a bug in the transformation.
I say 'semi-automated' because in cases where the tool can't figure out the problem, the tool could yield to human judgment. Tooling for this doesn't have to be perfect, but a small amount of automation could go a long way.
## Examples
example.ts
```ts
const myPromise: Promise<{}> = dontCarePromise();
```
tsc --upgrade --from=3.5 --to=3.6 example.ts
```ts
const myPromise: Promise<unknown> = dontCarePromise();
```
The example is from https://github.com/microsoft/TypeScript/issues/33272.
In a case where the conversion *would* lead to a new type error, I'd expect the tool to bail out, explain the situation, and let the human decide what to do.
## 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,Needs Proposal | low | Critical |
491,584,148 | kubernetes | Document that some specific properties may not be removed when using kubectl apply | Per #72519, serviceAccountName (and serviceAccount) are not removed when they are deleted from a resource spec. which is then passed to `kubectl apply`.
I hit this exact bug, and there is nothing in the docs at https://kubernetes.io/docs/concepts/cluster-administration/manage-deployment/#kubectl-apply to warn about the existence of special cases like that.
If there is no obvious fix that would address non-removal of special cases, then this caveat should be mentioned. | kind/documentation,sig/cli,sig/docs | medium | Critical |
491,587,181 | vue | `$forceUpdate` do not update checked status of checkbox- | ### Version
2.6.10
### Reproduction link
[https://jsfiddle.net/ahy27vu3/](https://jsfiddle.net/ahy27vu3/)
### Steps to reproduce
1. input[type="text"] value is forced to` ''`, which is controlled completely
2. input[type="checkbox"] checked does not work
### What is expected?
input[type="checkbox"] checked is forced to `false`
### What is actually happening?
input[type="checkbox"] checked keep the value last entered
---
Maybe this is designed on purpose，Could you please explain why？
https://github.com/vuejs/vue/blob/dev/src/platforms/web/runtime/modules/dom-props.js#L67
https://github.com/vuejs/vue/blob/dev/src/platforms/web/runtime/modules/dom-props.js#L42
<!-- generated by vue-issues. DO NOT REMOVE --> | regression | low | Major |
491,594,459 | flutter | webview_flutter plugin does not support "setNetworkAvailable" method from android webview | I have a flutter application that needs to display a web page that uses the "navigator.onLine" property. At the moment it only returns true no matter what. After some google searches it looks like this is a common issue with the android webview, and that there is a method you can call on the webview to change the property (https://stackoverflow.com/a/18746271/3407591). But this method does not exist for the webview_flutter plugin. Is there a another way for me to change this property in flutter or a way to call this function? | c: new feature,platform-android,d: stackoverflow,p: webview,package,P3,team-android,triaged-android | low | Minor |
491,682,224 | rust | `ManuallyDrop` should be a dyn-compatible receiver type | We should extend the set of library-defined self types on stable to include types involving `ManuallyDrop`. `ManuallyDrop` is a transparent wrapper around another type which is completely unconstrained; there should be no technical limitations on making it an dyn-compatible[^1] receiver type.
To be explicit, I am proposing to make the following types valid self types on stable, and all but the first would be dyn-compatible:
* `ManuallyDrop<Self>`
* `&ManuallyDrop<Self>`
* `&mut ManuallyDrop<Self>`
* `Box<ManuallyDrop<Self>>`
* `Rc<ManuallyDrop<Self>>`
* `Arc<ManuallyDrop<Self>>`
* `Pin<&ManuallyDrop<Self>>`
* `Pin<&mut ManuallyDrop<Self>>`
* `Pin<Box<ManuallyDrop<Self>>>`
* `Pin<Rc<ManuallyDrop<Self>>>`
* `Pin<Arc<ManuallyDrop<Self>>>`
This is similar to but distinct from #56193, which is about making coercions valid involving a manually drop wrapped around a pointer type.
In particular, I am personally interested in having `Pin<&mut ManuallyDrop<Self>>` as a dyn-compatible receiver type for an unsafe trait method in my code, as it assists users in understanding the guarantees the caller is giving them - that it will never access or move this value again, even to drop it, do with that what you will.
cc @mikeyhew @rust-lang/lang @rust-lang/wg-unsafe-code-guidelines
[^1]: Formerly known as "object safe". | T-lang,F-arbitrary_self_types,A-trait-objects | medium | Major |
491,733,105 | PowerToys | [FancyZones] Dynamic layouts bases off of focused Window | # Summary of the new feature/enhancement
At first I thought pressing windows + ~ selecting a new layout and clicking apply was too many steps. So I first thought a short cut key to change the layout quicker would be ideal. Then I realized, I wanted to change layouts based off of the focused window.
# Proposed technical implementation details
Looking at the default layouts there are Grids/Columns, and Priority Grid. A dynamic Priority Grid where the Priority can be any column would be the first step. IE: Priority Left, Right and Center. When focusing on the respective zone, the layout should change automatically.
| Idea-Enhancement,FancyZones-Layouts,Product-FancyZones | low | Minor |
491,751,667 | TypeScript | Declaration emit is broken for parameters marked /* @internal */ | .ts:
```ts
interface TypeChecker {
getContextualType(node: Expression, /* @internal */ contextFlags?: ContextFlags): Type | undefined;
}
```
.d.ts:
```ts
interface TypeChecker {
getContextualType(node: Expression,
: Type | undefined;
``` | Bug,Domain: Declaration Emit | low | Critical |
491,754,666 | youtube-dl | Request for site support Magellan TV | <!--
######################################################################
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 2019.09.01. 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 **2019.09.01**
- [ 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: https://www.magellantv.com/video/birth-of-planet-earth-4k
- Single video: https://www.magellantv.com/series/man-made-marvels-asia/chinas-forgotten-city
- Playlist: https://www.magellantv.com/series/man-made-marvels-asia
## Description
<!--
Provide any additional information.
If work on your issue requires account credentials please provide them or explain how one can obtain them.
-->
Free 1 month trial https://www.magellantv.com
| site-support-request | low | Critical |
491,765,055 | terminal | Bug: When headless, GetLargestConsoleWindowSize reports a MUCH too small size | <!--
🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨
I ACKNOWLEDGE THE FOLLOWING BEFORE PROCEEDING:
1. If I delete this entire template and go my own path, the core team may close my issue without further explanation or engagement.
2. If I list multiple bugs/concerns in this one issue, the core team may close my issue without further explanation or engagement.
3. If I write an issue that has many duplicates, the core team may close my issue without further explanation or engagement (and without necessarily spending time to find the exact duplicate ID number).
4. If I leave the title incomplete when filing the issue, the core team may close my issue without further explanation or engagement.
5. If I file something completely blank in the body, the core team may close my issue without further explanation or engagement.
All good? Then proceed!
-->
<!--
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
ssh client: recent internal build of Windows
ssh server: recent internal build of Windows (WCOS)
vim: custom build that can load on the server
```
In a headless pseudoconsole session (such as over ssh), `GetLargestConsoleWindowSize` is reporting a maximum possible window size of 640 columns and 18 rows (on my particular SKU of Windows, which might differ from others). 640 columns isn't bad, I suppose, but 18 rows is downright *tiny*. I debugged a bit and found that internally (in conhost) this is due to a "screen" size in pixels of 640x300, and a font size of {1, 16}. (Of course there is not really a screen or pixels in a headless scenario, and that font size seems strange.)
Over ssh, I believe the console does not have any way to determine the ssh client's maximum possible window size (it doesn't know the size of the monitor the client is using, or the font being used, or even if it's a windows machine). It can be notified of window size changes on the client side (section 6.7: "Window Dimension Change Message" of RFC 4254), but not what the max possible size is. (Although I know very little about ssh; it's possible there's some extension or convention or something.)
So I propose that since there is no monitor hooked up to a headless session, `GetLargestConsoleWindowSize` should just report `{ SHRT_MAX, SHRT_MAX }` (and let the client deal with whatever happens if the client sets the console's size too big).
(Without this, vim running in an ssh session is stuck at 18 lines high and cannot be made any bigger.)
| Help Wanted,Product-Conpty,Area-Server,Issue-Bug,Priority-2 | low | Critical |
491,796,907 | storybook | UI: Add external links to navigation | The left-hand navigation of storybook consists of links to stories, arranged hierarchically. The Storybook logo link is also configurable (as well as visually).
In 5.2, we added the ability to link to "documentation-only stories" which is markdown-based documentation that can be interspersed within your stories.
Let's also allow the user to embed links to other external resources in the same navigation tree.
## One possible implementation
In `storiesOf`:
```js
storiesOf(...)
.add('storybook website', 'http://storybook.js.org', { isLink: true })
```
In `CSF`:
```js
export default { ... }
export const storybookWebsite = 'http://storybook.js.org'
storybookWebsite.story {
isLink: true,
}
```
In `MDX`:
```jsx
<Navigation name='storybook website' href='http://storybook.js.org' />
```
| feature request,ui,csf | medium | Major |
491,852,182 | youtube-dl | please add support for quint.com | Dear Developers please add support for quint.com
```
$ youtube-dl --version
2019.09.01
```
```
$ youtube-dl -c -F --verbose "https://www.thequint.com/news/india/non-approved-company-private-engineers-checked-evm-in-elections"
[debug] System config: []
[debug] User config: []
[debug] Custom config: []
[debug] Command-line args: ['-c', '-F', '--verbose', 'https://www.thequint.com/news/india/non-approved-company-private-engineers-checked-evm-in-elections']
[debug] Encodings: locale UTF-8, fs utf-8, out UTF-8, pref UTF-8
[debug] youtube-dl version 2019.09.01
[debug] Python version 3.7.4+ (CPython) - Linux-5.2.0-2-amd64-x86_64-with-debian-bullseye-sid
[debug] exe versions: ffmpeg 4.1.4-1, ffprobe 4.1.4-1, phantomjs 2.1.1, rtmpdump 2.4
[debug] Proxy map: {}
[generic] non-approved-company-private-engineers-checked-evm-in-elections: Requesting header
WARNING: Falling back on generic information extractor.
[generic] non-approved-company-private-engineers-checked-evm-in-elections: Downloading webpage
[generic] non-approved-company-private-engineers-checked-evm-in-elections: Extracting information
ERROR: Unsupported URL: https://www.thequint.com/news/india/non-approved-company-private-engineers-checked-evm-in-elections
Traceback (most recent call last):
File "/usr/lib/python3/dist-packages/youtube_dl/YoutubeDL.py", line 796, in extract_info
ie_result = ie.extract(url)
File "/usr/lib/python3/dist-packages/youtube_dl/extractor/common.py", line 530, in extract
ie_result = self._real_extract(url)
File "/usr/lib/python3/dist-packages/youtube_dl/extractor/generic.py", line 3355, in _real_extract
raise UnsupportedError(url)
youtube_dl.utils.UnsupportedError: Unsupported URL: https://www.thequint.com/news/india/non-approved-company-private-engineers-checked-evm-in-elections
``` | site-support-request | low | Critical |
491,865,241 | TypeScript | Generators/Async Generators with TReturn type argument which extends undefined have a return method with a non-optional argument in 3.6.2 | <!-- 🚨 STOP 🚨 𝗦𝗧𝗢𝗣 🚨 𝑺𝑻𝑶𝑷 🚨
Half of all issues filed here are duplicates, answered in the FAQ, or not appropriate for the bug tracker. Even if you think you've found a *bug*, please read the FAQ first, especially the Common "Bugs" That Aren't Bugs section!
Please help us by doing the following steps before logging an issue:
* Search: https://github.com/Microsoft/TypeScript/search?type=Issues
* Read the FAQ: https://github.com/Microsoft/TypeScript/wiki/FAQ
Please fill in the *entire* template below.
-->
<!-- Please try to reproduce the issue with `typescript@next`. It may have already been fixed. -->
**TypeScript Version:** 3.6.2
<!-- Search terms you tried before logging this (so others can find this issue more easily) -->
**Search Terms:**
generator iterator return argument optional required
**Code**
```ts
function *gen(): Generator<number, number | undefined> {
yield 1;
return 2;
}
gen().return();
```
**Expected behavior:**
The argument passed to the `return` method should be optional. Not sure if this is possible with typescript.
**Actual behavior:**
```
file.ts(6,7): error TS2554: Expected 1 arguments, but got 0.
6 gen().return();
~~~~~~~~
typescript/lib/lib.es2015.generator.d.ts:26:12
26 return(value: TReturn): IteratorResult<T, TReturn>;
~~~~~~~~~~~~~~
An argument for 'value' was not provided.
```
**Playground Link:** <!-- A link to a TypeScript Playground "Share" link which demonstrates this behavior -->
Playground is not available for 3.6.2
**Related Issues:** <!-- Did you find other bugs that looked similar? -->
#30790 Strongly typed iterator PR.
#32682 Other (async) iterator usability issues.
#33239 Trouble implementing async iterator interface in 3.6.2
#33241 Unable to use done property of IteratorResult to narrow types in typescript 3.6.2 | Needs Investigation | low | Critical |
491,871,126 | pytorch | TestAutograd.test_deep_reentrant fails with SIGBUS on macOS | ## 🐛 Bug
TestAutograd.test_deep_reentrant fails with SIGBUS on macOS.
## To Reproduce
Steps to reproduce the behavior: run autograd tests on macOS.
## Environment
Found this while working on #25930.
cc @ezyang @SsnL @albanD @zou3519 | module: autograd,module: tests,triaged,module: macos | low | Critical |
491,877,744 | godot | Data races when running Godot | **Godot version:**
3.2.alpha.custom_build. 24e1039eb
**OS/device including version:**
Ubuntu 19.04
**Issue description:**
When I run Godot project manager, then thread sanitizer shows some data races(reading data in one thread, when saving in another thread)
- Log from project manager - [AAAA.txt](https://github.com/godotengine/godot/files/3597218/AAAA.txt)
```
WARNING: ThreadSanitizer: data race (pid=27914)
Atomic write of size 8 at 0x0000092a1ec0 by thread T1:
#0 __tsan_atomic64_fetch_add <null> (godot.x11.tools.64.llvms+0x1703076)
#1 unsigned long atomic_add<unsigned long, unsigned long>(unsigned long volatile*, unsigned long) /home/rafal/Pulpit/mojgodot/./core/safe_refcount.h:136:9 (godot.x11.tools.64.llvms+0x666442b)
#2 Memory::alloc_static(unsigned long, bool) /home/rafal/Pulpit/mojgodot/core/os/memory.cpp:98 (godot.x11.tools.64.llvms+0x666442b)
#3 operator new(unsigned long, char const*) /home/rafal/Pulpit/mojgodot/core/os/memory.cpp:42:9 (godot.x11.tools.64.llvms+0x6664268)
#4 ThreadPosix::thread_callback(void*) /home/rafal/Pulpit/mojgodot/drivers/unix/thread_posix.cpp:70:45 (godot.x11.tools.64.llvms+0x2d3e899)
Previous read of size 8 at 0x0000092a1ec0 by main thread:
#0 Memory::alloc_static(unsigned long, bool) /home/rafal/Pulpit/mojgodot/core/os/memory.cpp:99:42 (godot.x11.tools.64.llvms+0x6664441)
#1 CowData<wchar_t>::resize(int) /home/rafal/Pulpit/mojgodot/./core/cowdata.h:274:32 (godot.x11.tools.64.llvms+0x22ceb5b)
#2 String::resize(int) /home/rafal/Pulpit/mojgodot/./core/ustring.h:154:45 (godot.x11.tools.64.llvms+0x2efe9d4)
#3 String::copy_from(char const*) /home/rafal/Pulpit/mojgodot/core/ustring.cpp:169:2 (godot.x11.tools.64.llvms+0x64bc814)
#4 String::operator=(char const*) /home/rafal/Pulpit/mojgodot/core/ustring.cpp:331:2 (godot.x11.tools.64.llvms+0x64bd758)
#5 detect_prime() /home/rafal/Pulpit/mojgodot/platform/x11/detect_prime.cpp:135:15 (godot.x11.tools.64.llvms+0x1782c12)
#6 OS_X11::initialize(OS::VideoMode const&, int, int) /home/rafal/Pulpit/mojgodot/platform/x11/os_x11.cpp:275:16 (godot.x11.tools.64.llvms+0x174b148)
#7 Main::setup2(unsigned long) /home/rafal/Pulpit/mojgodot/main/main.cpp:1117:35 (godot.x11.tools.64.llvms+0x17bade5)
#8 Main::setup(char const*, int, char**, bool) /home/rafal/Pulpit/mojgodot/main/main.cpp:1062:10 (godot.x11.tools.64.llvms+0x17b9f28)
#9 main /home/rafal/Pulpit/mojgodot/platform/x11/godot_x11.cpp:49:14 (godot.x11.tools.64.llvms+0x1744730)
Location is global 'Memory::mem_usage' of size 8 at 0x0000092a1ec0 (godot.x11.tools.64.llvms+0x0000092a1ec0)
Thread T1 (tid=27916, running) created by main thread at:
#0 pthread_create <null> (godot.x11.tools.64.llvms+0x16b90d5)
#1 ThreadPosix::create_func_posix(void (*)(void*), void*, Thread::Settings const&) /home/rafal/Pulpit/mojgodot/drivers/unix/thread_posix.cpp:90:2 (godot.x11.tools.64.llvms+0x2d3eadf)
#2 Thread::create(void (*)(void*), void*, Thread::Settings const&) /home/rafal/Pulpit/mojgodot/core/os/thread.cpp:51:10 (godot.x11.tools.64.llvms+0x666c8b0)
#3 IP::IP() /home/rafal/Pulpit/mojgodot/core/io/ip.cpp:326:22 (godot.x11.tools.64.llvms+0x679c0df)
#4 IP_Unix::IP_Unix() /home/rafal/Pulpit/mojgodot/drivers/unix/ip_unix.cpp:265:10 (godot.x11.tools.64.llvms+0x2f208dc)
#5 IP_Unix::_create_unix() /home/rafal/Pulpit/mojgodot/drivers/unix/ip_unix.cpp:262:9 (godot.x11.tools.64.llvms+0x2f2085b)
#6 IP::create() /home/rafal/Pulpit/mojgodot/core/io/ip.cpp:310:9 (godot.x11.tools.64.llvms+0x679bde3)
#7 register_core_types() /home/rafal/Pulpit/mojgodot/core/register_core_types.cpp:211:7 (godot.x11.tools.64.llvms+0x641b468)
#8 Main::setup(char const*, int, char**, bool) /home/rafal/Pulpit/mojgodot/main/main.cpp:341:2 (godot.x11.tools.64.llvms+0x17abf12)
#9 main /home/rafal/Pulpit/mojgodot/platform/x11/godot_x11.cpp:49:14 (godot.x11.tools.64.llvms+0x1744730)
SUMMARY: ThreadSanitizer: data race (/home/rafal/Pulpit/mojgodot/bin/godot.x11.tools.64.llvms+0x1703076) in __tsan_atomic64_fetch_add
```
**Steps to reproduce:**
1. Compile Godot with thread sanitizer support e.g.(newer versions of GCC compiler are a little buggy for now, so llvm is the only option)
```
scons p=x11 -j6 use_tsan=yes use_llvm=yes
```
2. Run project manager
```
bin/godot.x11.tools.64.llvms
```
| bug,topic:core | low | Critical |
491,879,322 | flutter | Integrate ./flutter/ci/build_flutter_runner_tests.sh with ./flutter/testing/run_tests.py. | That is the primary entrypoint for all testing. Currently, the bash script also depends on only the x86 Fuchsia variant being built. | customer: fuchsia,engine,platform-fuchsia,P2,team-engine,triaged-engine | low | Minor |
491,883,985 | go | cmd/go: upgrading a module gives different results based on "how" you upgrade | ### Summary
From the Go Wiki:
>Day-to-day upgrading and downgrading of dependencies should be done using 'go get', which will automatically update the go.mod file. Alternatively, you can edit go.mod directly.
From `go help modules`
> If go.mod is edited directly, commands
like 'go build' or 'go list' will assume that an upgrade is intended and
automatically make any implied upgrades and update go.mod to reflect them.
Therefore, people can upgrade a module by:
1. Simply running `go get <mod>@<newVersion>` **or**
2. Directly editing the go.mod file and replace the old version with the new one.
Most Go developers I've talked to, including myself, have believed that both options are interchangeable. However, whether a user picks option 1 or option 2 above can actually lead to a different final build list.
While I'm not sure this is a bug, but the fact that *how* you upgrade a module can lead to different results seems odd or at least something worth documenting.
### Reproduce:
I have replicated the same dependency tree in the MVS article here: https://research.swtch.com/vgo-mvs
You can find that dependency tree under https://github.com/gomvs
github.com/gomvs/a is the main module and its current commit resides at Algorithm 1:
```
module github.com/gomvs/a
go 1.13
require (
github.com/gomvs/b v1.2.0
github.com/gomvs/c v1.2.0
)
```
If we want to trigger Algorithm 3 by upgrading one module (c v1.2.0 to c v1.3.0) we have 2 ways of doing it:
#### Option 1: `go get github.com/gomvs/[email protected]`
This will properly implement Algorithm R by adding Module D v1.4.0 to the go.mod file because the Rough List still included [email protected]
Therefore, the upgraded go.mod file will look like this:
```
module github.com/gomvs/a
go 1.13
require (
github.com/gomvs/b v1.2.0
github.com/gomvs/c v1.3.0
github.com/gomvs/d v1.4.0 // indirect
)
```
If you run `go run .` you'll see the following dependency tree being called:
```
A
B v1.2.0
D v1.4.0
E v1.2.0
C v1.3.0
F v1.1.0
G v1.1.0
```
#### Option 2: Directly edit the go.mod file by going to line 7 in go.mod and simply changing `v1.2.0` to `v1.3.0`
If a user decided to upgrade from c v1.2.0 to c v1.3.0 by "editing the go.mod file directly" then Algorithm R has no way of remembering that we had c v1.2.0 when running "go build" and therefore we end up downgrading to Module D v1.3.0.
The resulting go.mod file remains the same:
```
module github.com/gomvs/a
go 1.13
require (
github.com/gomvs/b v1.2.0
github.com/gomvs/c v1.3.0
)
```
And if you run `go run .` you get the following results:
```
A
B v1.2.0
D v1.3.0
E v1.2.0
C v1.3.0
F v1.1.0
G v1.1.0
```
Notice D v1.3.0 (and not D v1.4.0)
#### The MVS article mentions this behavior as "incorrect" and therefore should we consider directly upgrading the go.mod file incorrect? Also, should we document this behavior?
There is actually option 3, which I didn't know we could do:
#### Option 3: duplicate modules in go.mod file
You can actually have a go.mod file that looks like this:
```
module github.com/gomvs/a
go 1.13
require (
github.com/gomvs/b v1.2.0
github.com/gomvs/c v1.2.0
github.com/gomvs/c v1.3.0
)
```
Notice, that `c` is mentioned twice at two different versions. Running `go mod tidy` ends up picking the correct version, and also adding `d` at v1.4.0 correctly. Given Algorithm R, this makes sense and I wonder if we should also mention it?
cc: @bcmills | Documentation,help wanted,NeedsInvestigation,modules | low | Critical |
491,929,596 | flutter | Better document the CHROME_EXECUTABLE environment variable | Hello, there seems to not be much documentation on CHROME_EXECUTABLE available.
There's info on https://github.com/flutter/flutter/wiki/Building-a-web-application-with-Flutter but it is otherwise pretty hidden.
It could be shown on the Web FAQ, for example. | d: api docs,platform-web,P3,team-web,triaged-web | low | Minor |
491,945,559 | opencv | WASM Face Module and DLib Facial Landmarks JavaScript Build | 1. How can we include the face module to detect real-time facial landmarks to the JavaScript build because it is also creating errors and I am not sure where the bug occurs because I followed the same instructions from the website? In fact, I tried FacemarkKazemi, FacemarkAAM, and FacemarkLBF.
2. How can we include the dlib's facial landmark model library into the OpenCV.js build because there are dependencies?
| feature,category: contrib,category: javascript (js) | low | Critical |
491,994,510 | flutter | The Fuchsia Toolchain & SDK Autoroller should use the revision tags instead of instance IDs. | Currently, we have 4 Fuchsia specific autorollers. One for the toolchain and another for the SDK on Linux & Mac. Each autoroller uses the instance ID of CIPD packages. This is disadvantageous because the versions of the toolchain and SDK on Mac and Linux may not always be in sync with each other. Also, since there are 4 instance fields to update, there are 4 rollers.
Instead of instance IDs, the autorollers should use a tag for the SDK as well as the toolchain. The tag (int the `git:<sha>` format) for both the Linux and Mac variants of the toolchain and SDK should the same. This guarantees that host artifacts built on the Mac will not drift from those built on Linux. An added benefit is that we will require just two autorollers instead of 4. One for the SDK and another for the roller. | team,customer: fuchsia,engine,dependency: fuchsia,P3,team-engine,triaged-engine | low | Major |
491,998,363 | TypeScript | support destructuring parameter @param tag in tsdoc |
tsdoc rule:
https://github.com/microsoft/tsdoc/issues/186
vscode issue:
https://github.com/microsoft/vscode/issues/80368
example:
```ts
/**
*
* @param bar - the bar parameter
* @param a - the first parameter
* @param b - the second parameter
* @param c - the third parameter
* @returns the return value
*/
function foo(bar: string, {
a,
b,
c
}:{
a: string,
b: number,
c: string
}) {
return bar + a;
}
```
| Suggestion,Experience Enhancement | low | Major |
492,018,796 | pytorch | Provide a way to select SVD algorithm in PyTorch? | PyTorch seems to use `gesdd` for singular value decomposition. Alternative `gesvd` is slower, but more accurate/robust. PyTorch could follow scipy syntax and provide a way to choose between them using `lapack_driver` kwarg
Some background on gesvd vs gesdd
https://epubs.siam.org/doi/10.1137/17M1117732
https://savannah.gnu.org/bugs/?55564
```
import numpy as np
import torch, scipy
fn = "pinv2_crash.txt"
if not os.path.exists(fn):
import urllib.request
url="https://s3.amazonaws.com/yaroslavvb2/data/"+fn
response = urllib.request.urlopen(url)
body = response.read()
print("Read %d bytes"%(len(body),))
open(fn, "wb").write(body)
mat = np.genfromtxt(fn, delimiter= ",").astype(np.float32)
scipy.linalg.svd(mat, lapack_driver="gesdd") # crashes
torch.svd(torch.tensor(mat)) # crashes
scipy.linalg.svd(mat, lapack_driver="gesvd") # works
```
cc @vincentqb @vishwakftw @jianyuh @nikitaved @pearu @mruberry @heitorschueroff @SsnL | triaged,module: linear algebra,function request | low | Critical |
492,020,701 | pytorch | [C++] `Module::pretty_print` is broken | ## 🐛 Bug
<!-- A clear and concise description of what the bug is. -->
As described in the documentation, `Module::pretty_print` prints the name of the current module and prints its submodules recursively.
> Streams a pretty representation of the Module into the given stream.
>
> By default, this representation will be the name of the module (taken from name()), followed by a recursive pretty print of all of the Module’s submodules.
>
> Override this method to change the pretty print. The input stream should be returned from the method, to allow easy chaining.
However, in reality, `pretty_print` function merely prints the name of the current module. `pretty_print_recursive` will print its submodules recursively, but it's ~~not virtual~~ private meaning it's not inherited by subclasses. Classes like `Sequence` even don't have such a method.
## To Reproduce
<!-- 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. -->
Make `pretty_print` correctly print the module hierarchy.
## Environment
- PyTorch Version (e.g., 1.0): Nightly
- OS (e.g., Linux): Linux
- How you installed PyTorch (`conda`, `pip`, source):
- Build command you used (if compiling from source):
- Python version:
- CUDA/cuDNN version:
- GPU models and configuration:
- Any other relevant information:
## Additional context
<!-- Add any other context about the problem here. -->
cc @yf225 | module: cpp,module: nn,triaged | low | Critical |
492,029,990 | go | x/playground: support for syntax highlighting | Currently, the playground is yellow background and black text, make it difficult to read the codes. Can you make codes colorful and bring editor to playground (like [Wandbox](https://wandbox.org/)) ? | NeedsInvestigation,FeatureRequest | low | Major |
492,068,455 | ant-design | Vertical progress bar | - [x] I have searched the [issues](https://github.com/ant-design/ant-design/issues) of this repository and believe that this is not a duplicate.
### What problem does this feature solve?
need of vertical progress bar
### What does the proposed API look like?
have something like a "orientation" prop for the progress bar
<!-- generated by ant-design-issue-helper. DO NOT REMOVE --> | Inactive | low | Minor |
492,070,407 | flutter | RefreshIndicator, show without launch onRefresh | Is there any way to achieve this behavior? | c: new feature,framework,f: material design,d: stackoverflow,P3,team-design,triaged-design | low | Major |
492,071,101 | rust | closure compilation fails depending on type inference | ```rust
fn foo(_: &mut String) {}
fn bug1() {
let _ = |a: &mut String| for _ in 0..0 {
foo(a)
}; //works fine
let _ = |a| for _ in 0..0 {
foo(a)
}; //fails
}
fn bug2() {
let mut a = String::new();
let mut b = String::new();
let f = |a: &mut String| foo(a);
let g = |b| foo(b);
f(&mut a); //works
f(&mut a); //works
g(&mut b); //works
g(&mut b); //fails
}
```
([Playground](https://play.rust-lang.org/?version=stable&mode=debug&edition=2018&gist=772e2a833bd2bdcf4296d3667e2ff4b7))
Errors:
```
Compiling playground v0.0.1 (/playground)
error[E0382]: use of moved value: `a`
--> src/lib.rs:8:13
|
7 | let _ = |a| for _ in 0..0 {
| - move occurs because `a` has type `&mut std::string::String`, which does not implement the `Copy` trait
8 | foo(a)
| ^ value moved here, in previous iteration of loop
error[E0499]: cannot borrow `b` as mutable more than once at a time
--> src/lib.rs:20:7
|
19 | g(&mut b); //works
| ------ first mutable borrow occurs here
20 | g(&mut b); //fails
| - ^^^^^^ second mutable borrow occurs here
| |
| first borrow later used by call
error: aborting due to 2 previous errors
Some errors have detailed explanations: E0382, E0499.
For more information about an error, try `rustc --explain E0382`.
error: Could not compile `playground`.
To learn more, run the command again with --verbose.
``` | A-closures,A-inference | low | Critical |
492,165,698 | go | cmd/go: list -compiled fails to populate CompiledGoFiles when the resulting package cannot be linked | <!-- Please answer these questions before submitting your issue. Thanks! -->
### What version of Go are you using (`go version`)?
<pre>
$ go version
go version go1.13 linux/amd64
</pre>
### Does this issue reproduce with the latest release?
Yes.
### What operating system and processor architecture are you using (`go env`)?
Linux, amd64
### What did you do?
Create a package with the following contents:
```
package pkg
// #cgo LDFLAGS: -ldoesnotexist
import "C"
```
Run `go list -json -x -compiled` on the package
### What did you expect to see?
No errors, no attempted linking of object files, complete JSON output (meaning CompiledGoFiles is populated).
### What did you see instead?
```
$ go list -e -json -compiled -x
WORK=/tmp/go-build309865873
mkdir -p $WORK/b001/
cd /home/dominikh/prj/src/sandbox/bar
CGO_LDFLAGS='"-g" "-O2" "-ldoesnotexist"' /usr/lib/go/pkg/tool/linux_amd64/cgo -objdir $WORK/b001/ -importpath sandbox/bar -- -I $WORK/b001/ -g -O2 ./bar.go
cd $WORK
gcc -fno-caret-diagnostics -c -x c - -o /dev/null || true
gcc -Qunused-arguments -c -x c - -o /dev/null || true
gcc -fdebug-prefix-map=a=b -c -x c - -o /dev/null || true
gcc -gno-record-gcc-switches -c -x c - -o /dev/null || true
cd $WORK/b001
TERM='dumb' gcc -I /home/dominikh/prj/src/sandbox/bar -fPIC -m64 -pthread -fmessage-length=0 -fdebug-prefix-map=$WORK/b001=/tmp/go-build -gno-record-gcc-switches -I ./ -g -O2 -o ./_x001.o -c _cgo_export.c
TERM='dumb' gcc -I /home/dominikh/prj/src/sandbox/bar -fPIC -m64 -pthread -fmessage-length=0 -fdebug-prefix-map=$WORK/b001=/tmp/go-build -gno-record-gcc-switches -I ./ -g -O2 -o ./_x002.o -c bar.cgo2.c
TERM='dumb' gcc -I /home/dominikh/prj/src/sandbox/bar -fPIC -m64 -pthread -fmessage-length=0 -fdebug-prefix-map=$WORK/b001=/tmp/go-build -gno-record-gcc-switches -I ./ -g -O2 -o ./_cgo_main.o -c _cgo_main.c
cd /home/dominikh/prj/src/sandbox/bar
TERM='dumb' gcc -I . -fPIC -m64 -pthread -fmessage-length=0 -fdebug-prefix-map=$WORK/b001=/tmp/go-build -gno-record-gcc-switches -o $WORK/b001/_cgo_.o $WORK/b001/_cgo_main.o $WORK/b001/_x001.o $WORK/b001/_x002.o -g -O2 -ldoesnotexist
# sandbox/bar
/usr/bin/ld: cannot find -ldoesnotexist
collect2: error: ld returned 1 exit status
{
"Dir": "/home/dominikh/prj/src/sandbox/bar",
"ImportPath": "sandbox/bar",
"Name": "pkg",
"Target": "/home/dominikh/prj/pkg/linux_amd64/sandbox/bar.a",
"Root": "/home/dominikh/prj/",
"Match": [
"."
],
"Stale": true,
"StaleReason": "build ID mismatch",
"CgoFiles": [
"bar.go"
],
"CgoLDFLAGS": [
"-ldoesnotexist"
],
"Imports": [
"C",
"unsafe",
"runtime/cgo",
"syscall"
],
"Deps": [
"errors",
"internal/bytealg",
"internal/cpu",
"internal/oserror",
"internal/race",
"internal/reflectlite",
"runtime",
"runtime/cgo",
"runtime/internal/atomic",
"runtime/internal/math",
"runtime/internal/sys",
"sync",
"sync/atomic",
"syscall",
"unsafe"
]
}
```
That is, Go actually attempts to build the package and fails due to the missing library we're trying to link against. It also doesn't populate CompiledGoFiles, even though cgo preprocessing succeeded.
As I understand it, this is trying to do more work than is necessary. It also unnecessarily fails.
The `-compiled` flag is documented as follows:
> The -compiled flag causes list to set CompiledGoFiles to the Go source
> files presented to the compiler. Typically this means that it repeats
> the files listed in GoFiles and then also adds the Go code generated
> by processing CgoFiles and SwigFiles. The Imports list contains the
> union of all imports from both GoFiles and CompiledGoFiles.
It does not imply actual compilation. Also note that we're not passing the `-exported` flag. Furthermore, in order to do cgo processing, we do not need the linker. For that, looking at headers suffices, and `go tool cgo foo.go` does succeed.
After a cursory look at cmd/go/internal/work/exec.go, (*Builder).build, it seems that if we require CompiledGoFiles, and they can't be found in the cache, we execute the full cgo pipeline used for building. In theory, it should be possible to optimize this.
The current behavior is problematic for two reasons, from the point of view of tools that analyze Go code:
1. it is slower than necessary. Analyzing a large cgo-using program will be significantly faster if we can skip the linking phase.
2. it breaks certain setups of running static analysis. For example, a test environment may have the necessary headers available, but no object files for the libraries. We should be able to do cgo preprocessing on the code, and thus be able to do our static analysis. Requiring the linker to run successfully prevents this from happening.
/cc @matloob @ianthehat @bcmills | help wanted,NeedsInvestigation,GoCommand | low | Critical |
492,187,284 | go | runtime: futex contention caused by memory allocation in goroutines | <!-- Please answer these questions before submitting your issue. Thanks! -->
### What version of Go are you using (`go version`)?
Go 1.13
<pre>
$ go version
go version go1.13 linux/amd64
</pre>
### Does this issue reproduce with the latest release?
I made similar obsverations with older Go versions (go 1.10.x,, go1.11.x, and go1.12.x).
### What operating system and processor architecture are you using (`go env`)?
System: Ubuntu 18.04.3 LTS
Kernel: Linux 4.15.0-60-generic
Processor: Intel(R) Xeon(R) CPU E3-1230 V2 @ 3.30GHz (8 cores)
<details><summary><code>go env</code> Output</summary><br><pre>
$ go env
GO111MODULE=""
GOARCH="amd64"
GOBIN=""
GOCACHE="/home/felix/.cache/go-build"
GOENV="/home/felix/.config/go/env"
GOEXE=""
GOFLAGS=""
GOHOSTARCH="amd64"
GOHOSTOS="linux"
GONOPROXY=""
GONOSUMDB=""
GOOS="linux"
GOPATH="/home/felix/.gvm/pkgsets/go1.13/global:/home/felix/regmatch"
GOPRIVATE=""
GOPROXY="https://proxy.golang.org,direct"
GOROOT="/home/felix/.gvm/gos/go1.13"
GOSUMDB="sum.golang.org"
GOTMPDIR=""
GOTOOLDIR="/home/felix/.gvm/gos/go1.13/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-build193330363=/tmp/go-build -gno-record-gcc-switches"
</pre></details>
### What did you do?
Iteratively solving two tasks in parallel on a multi-core computer in two separate goroutines. The tasks are independent of each other. At the end of an iteration, the tasks' results are combined. Both tasks have similar running times, but they are not identical. Both tasks allocate local memory (this seems to be crucial for the problem!).
Whether using a wait group or channels at the end of an iteration does not seem to matter. The code below uses a wait group. See also
[main.txt](https://github.com/golang/go/files/3600480/main.txt).
* The main() function comprises a for loop with the iteration in which the tasks solved in each iteration.
* The two tasks are performed by the task() function.
* The main() function has also some flags, e.g., for setting the number of OS threads and for CPU profiling.
```
package main
import (
"flag"
"fmt"
"math/rand"
"os"
"runtime"
"runtime/pprof"
"sync"
"time"
)
func main() {
profile := flag.Bool("profile", false, "cpu profile")
cpu := flag.Int("cpu", runtime.GOMAXPROCS(-1), "number of OS threads")
iterations := flag.Int("num", 100000, "number of iterations")
//tasks := flag.Int("tasks", 2, "number tasks per iteration")
work := flag.Int("work", 20, "work in each task")
flag.Parse()
if *profile {
var f *os.File
var err error
if f, err = os.Create(fmt.Sprintf("cpu%d.prof", *cpu)); err != nil {
fmt.Printf("Failed to create file for CPU profiling (%s).\n", err)
return
}
if err = pprof.StartCPUProfile(f); err != nil {
fmt.Printf("Failed to start CPU profile (%s).\n", err)
return
}
defer func() {
pprof.StopCPUProfile()
if err = f.Close(); err != nil {
fmt.Printf("Failed to close file for CPU profiling (%s).\n", err)
}
}()
}
runtime.GOMAXPROCS(*cpu)
fmt.Printf("Using %d OS threads...\n", runtime.GOMAXPROCS(-1))
for i := 0; i < *iterations; i++ {
n := *work - rand.Intn(8)
in0, in1 := n+rand.Intn(4), n+rand.Intn(4)
var out0, out1 int
var dur0, dur1 time.Duration
var wg sync.WaitGroup
wg.Add(2)
go task(in0, &out0, &dur0, &wg)
go task(in1, &out1, &dur1, &wg)
wg.Wait()
fmt.Printf("%6d: %10s %10s %10s \t %d, %d\n", i, dur0, dur1, dur0-dur1, out0, out1)
}
}
func task(in int, res *int, dur *time.Duration, wg *sync.WaitGroup) {
stopwatch := time.Now()
memory := make([]int, fib(in))
for j := 0; j < len(memory); j++ {
memory[j] = j
}
*res = memory[len(memory)-1]
// Allocating some memory in the groroutine seems crucial for futex contention.
// The bad performance behavior is not observable for *res = fib(n).
*dur = time.Since(stopwatch)
wg.Done()
}
func fib(n int) int {
switch n {
case 0:
return 0
case 1:
return 1
default:
return fib(n-1) + fib(n-2)
}
}
```
### What did you expect to see?
I expect that the program runs faster on >=2 CPU cores than on a single CPU core, since the tasks are carried out in parallel. I also expect that the running times are similar on 2 and 8 CPU cores. There might be some scheduling overhead when using 8 CPU cores.
Note that -work flag allows you to control the hardness of the problem in the iteration. This parameter will depend on the computer you are using. The default parameters worked well on my computer for the issue.
### What did you see instead?
The program runs indead faster with 2 CPU cores than with a single CPU core. However, on 8 CPU cores it runs significantly slower (9 seconds compared to 13 seconds; the system time as output by /usr/bin/time also increases significantly). See the file
[output.txt](https://github.com/golang/go/files/3600469/output.txt)/
for a summary of multiple runs.
Looking at the CPU profiles, one see a significant increase by runtime.futex. It seems that the goroutines are interrupted by a futex syscall, which originates from the memory allocation in one of the goroutines. Afterwards, the Go scheduler then tries to find a runnable goroutine. However, I am not very familiar with futexes and the Go scheduler. So, I might be completely wrong here.
Overall, I would have expected a far smaller scheduling overhead and unused resources (i.e., CPU cores) should not slow down the program. | NeedsInvestigation,compiler/runtime | low | Critical |
492,212,099 | pytorch | Why doc building isn't failing us for referring to a non-existent method? | ## 📚 Documentation
<!-- A clear and concise description of what content in https://pytorch.org/docs is an issue. If this has to do with the general https://pytorch.org website, please file an issue at https://github.com/pytorch/pytorch.github.io/issues/new/choose instead. If this has to do with https://pytorch.org/tutorials, please file an issue at https://github.com/pytorch/tutorials/issues/new -->
Based on #25706, [toDense](https://pytorch.org/docs/stable/sparse.html#torch.sparse.FloatTensor.toDense) was a typo for [to_dense](https://github.com/pytorch/pytorch/blob/5d3267cd303817e0097a833bbd3f59be35295441/aten/src/ATen/core/TensorBody.h#L702) method but the doc building did not fail.
cc: @ezyang | module: docs,triaged | low | Minor |
492,274,518 | PowerToys | Some sort of staggered startup tool | # Summary of the new feature/enhancement
Like many users, I have a number of apps that automatically start up and on several of my machines, I have multiple online drives that start (Google Drive, DropBox, etc.) and when the system boots, all these apps are competing for resources, but aren't absolutely necessary _immediately_. The idea here is to be able to create a list of the automatic startup apps, etc. and be able to prioritize them so that they start up in sequence, and don't kill the system trying to compete with each other.
<!--
A clear and concise description of what the problem is that the new feature would solve.
Describe why and how a user would use this new functionality (if applicable).
-->
Ideally, the proposed tool would create and maintain an internal order of autorun apps, and it would be able to have multiple items at the same priority level. Each "level" would run in parallel, and the items at that "level" would get started and "settle down" before moving to the next level. The normal resource constraints I see are usually disk & cpu. (Even on an SSD this can get overwhelming).
As part of the implementation, the tool would look at existing autoruns (like the SysInternals utility) and watch those locations, and if something gets added automatically, it would strip it and add it to it's own list and notify the user (probably by balloon tip/system toast) that it had been added to its list and added to the "bottom" of the list.
# Proposed technical implementation details (optional)
1. The tool would watch the typical autorun locations - not to the depth of SysInternals (as a user, I don't care about automatic drivers, etc.)
2. It would maintain an icon in the system tray
3. The system tray would allow you to right click and view the settings
4. The system tray would also watch the autorun locations, and if they were modified it would provide a balloon tip/toast message to that fact
5. The system tray would allow the user to right click and edit the "levels"
6,. The number of "levels" would be arbitrary (although I can't see where people would need more than 255, and even that's excessive)
7. All the apps in a given "level" would be kicked off simultaneously
8. The user could arbitrarily add items to the "levels"
9. The apps in the various "levels" could be moved by drag & drop, including creating a new "level"; but it would not create empty levels
<!--
A clear and concise description of what you want to happen.
-->
| Idea-New PowerToy | low | Major |
492,281,075 | electron | BrowserWindow: missing methods for retrieving set options | <!-- As an open source project with a dedicated but small maintainer team, it can sometimes take a long time for issues to be addressed so please be patient and we will get back to you as soon as we can.
-->
### Preflight Checklist
<!-- Please ensure you've completed the following steps by replacing [ ] with [x]-->
* [x] I have read the [Contributing Guidelines](https://github.com/electron/electron/blob/master/CONTRIBUTING.md) for this project.
* [x] I agree to follow the [Code of Conduct](https://github.com/electron/electron/blob/master/CODE_OF_CONDUCT.md) that this project adheres to.
* [x] I have searched the issue tracker for a feature request that matches the one I want to file, without success.
The following options that can be passed to `BrowserWindow`'s constructor can be both retrieved and set even after instantiation:
- `alwaysOnTop`
- `autoHideMenuBar`
- `fullscreen`
- `simpleFullscreen`
- `hasShadow`
- `kiosk`
- `opacity`
- `parent`
- `title`
- `closable`
- `fullscreenable`
- `maximizable`
- `minimizable`
- `movable`
- `resizable`
- `x`
- `y`
- `width`
- `height`
- `minWidth`
- `minHeight`
- `maxWidth`
- `maxHeight`
But the following ones can only be set, there are no methods for retrieving their current values:
- `autoHideCursor`
- `backgroundColor`
- `skipTaskbar`
- `vibrancy`
I think methods for retrieving those values should be added for two reasons:
- Completeness: currently not all options that can be set at runtime can be also retrieved.
- I'm writing a couple of functions, one that retrieves an object of currently set, and configurable, options, another one for updating them. My use case is: retrieving the current options, updating the options a few times, restore the previous options, I can't really do it as I should be able to because I can't retrieve the value of those 4 properties I listed. | enhancement :sparkles:,component/BrowserWindow | low | Minor |
492,333,077 | flutter | Failed assertion: line 1773 pos 12: '_elements.contains(element)': is not true. | Hi,
I was just playing a bit around to test the new structured errors.
So I added a `SingleChildScrollView`around the column of the default counter app and placed a Expanded around the Column which lead to an exception. I removed the Expanded again and saved and got this error here.
```
════════ Exception caught by widgets library ═══════════════════════════════════
The following assertion was thrown building Listener:
'package:flutter/src/widgets/framework.dart': Failed assertion: line 1773 pos 12: '_elements.contains(element)': is not true.
Either the assertion indicates an error in the framework itself, or we should provide substantially more information in this error message to help you determine and fix the underlying cause.
In either case, please report this assertion by filing a bug on GitHub:
https://github.com/flutter/flutter/issues/new?template=BUG.md
User-created ancestor of the error-causing widget was
SingleChildScrollView
lib\main.dart:70
When the exception was thrown, this was the stack
#2 _InactiveElements.remove
package:flutter/…/widgets/framework.dart:1773
#3 Element._retakeInactiveElement
package:flutter/…/widgets/framework.dart:3059
#4 Element.inflateWidget
package:flutter/…/widgets/framework.dart:3082
#5 Element.updateChild
package:flutter/…/widgets/framework.dart:2897
#6 SingleChildRenderObjectElement.mount
package:flutter/…/widgets/framework.dart:5129
```
This is the code while the error happens:
```
return Scaffold(
appBar: AppBar(
title: Text(widget.title),
),
body: Center(
child: SingleChildScrollView(
child: Column(
mainAxisAlignment: MainAxisAlignment.center,
children: <Widget>[
Text(
'You have pushed the button this many times:',
),
Text(
'$_counter',
style: Theme.of(context).textTheme.display1,
),
],
),
),
),
floatingActionButton: FloatingActionButton(
onPressed: _incrementCounter,
tooltip: 'Increment',
child: Icon(Icons.add),
),
);
}
}
```
Removing the `SingleChildScrollView` and adding it again clears the exception.
I use VS code with;
[√] Flutter (Channel master, v1.10.1-pre.41, on Microsoft Windows [Version 10.0.18362.295], locale de-DE)
• Flutter version 1.10.1-pre.41 at C:\Entwicklung\flutter
• Framework revision 5a3a46ada2 (4 days ago), 2019-09-07 18:37:41 -0400
• Engine revision c9ea4dba8d
• Dart version 2.5.0 (build 2.5.0-dev.4.0 ec7ec4ecf7)
[√] Android toolchain - develop for Android devices (Android SDK version 29.0.2)
• Android SDK at C:\Users\escam\AppData\Local\Android\sdk
• Android NDK location not configured (optional; useful for native profiling support)
• 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.
[√] 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)
[√] IntelliJ IDEA Community Edition (version 2018.2)
• IntelliJ at C:\Program Files\JetBrains\IntelliJ IDEA Community Edition 2018.2.5
• Flutter plugin version 29.1.3
• Dart plugin version 182.4892.25
[√] VS Code, 64-bit edition (version 1.38.0)
• VS Code at C:\Program Files\Microsoft VS Code
• Flutter extension version 3.4.1
[√] Connected device (1 available)
• Android SDK built for x86 64 • emulator-5554 • android-x64 • Android 7.1.1 (API 25) (emulator)
| framework,a: error message,P2,team-framework,triaged-framework | low | Critical |
492,333,189 | flutter | [google_maps_flutter] Improve flutter implementation of google maps api | ## Use case
Hi, i'm currently working on an app, and there is some features i would like to implement, like:
- Use methods `showInfoWindow` and `hideInfoWindow`of `Marker`
- Use `ImageConfiguration` to set marker size (https://github.com/flutter/flutter/issues/34657)
- Use a custom `InfoWindow` on `Marker`
But i saw @amirh mentioning on an issue he was working on changing the maps plugin to a object based instead of controller based implementation, unfortunately i couldn't find that issue again.
## Proposal
I started doing `showInfoWindow` and `hideInfoWindow` inside `GoogleMapController` but i don't think it's the best way to do it. Maybe each marker should have it's `showInfoWindow` method, but should i implement a `MethodChannel` only for markers? I'm openning this issue to discuss it and would appreciate some instructions on how to do it. Sorry if it's the wrong place for it.
| c: new feature,a: quality,p: maps,package,c: proposal,team-ecosystem,P3,triaged-ecosystem | low | Minor |
492,358,106 | PowerToys | Show folder size in explorer | # Summary of the new feature/enhancement
Add the sum of the size of all files contained in the size file zone of the file explorer.
# Proposed technical implementation details (optional)
<!--
A clear and concise description of what you want to happen.
-->
| Idea-New PowerToy,Product-Tweak UI Design,Product-File Explorer | high | Critical |
492,415,013 | material-ui | [Chip] Add group component | <!-- Provide a general summary of the feature in the Title above -->
<!--
Thank you very much for contributing to Material-UI by creating an issue! ❤️
To avoid duplicate issues we ask you to check off the following list.
-->
<!-- Checked checkbox should look like this: [x] -->
- [x] I have searched the [issues](https://github.com/mui-org/material-ui/issues) of this repository and believe that this is not a duplicate.
## Summary 💡
I read the documentation for using material design Chip component. but I found that there is no react component for them. please add these features.
thank in advance.
<!-- Describe how it should work. -->
## Examples 🌈
<!--
Provide a link to the Material design specification, other implementations,
or screenshots of the expected behavior.
-->
## Motivation 🔦
<!--
What are you trying to accomplish? How has the lack of this feature affected you?
Providing context helps us come up with a solution that is most useful in the real world.
-->
| new feature,waiting for 👍,component: chip | low | Major |
492,432,953 | TypeScript | 3.6 regression: unions of callable types involving `this` types | <!-- 🚨 STOP 🚨 𝗦𝗧𝗢𝗣 🚨 𝑺𝑻𝑶𝑷 🚨
Half of all issues filed here are duplicates, answered in the FAQ, or not appropriate for the bug tracker. Even if you think you've found a *bug*, please read the FAQ first, especially the Common "Bugs" That Aren't Bugs section!
Please help us by doing the following steps before logging an issue:
* Search: https://github.com/Microsoft/TypeScript/search?type=Issues
* Read the FAQ: https://github.com/Microsoft/TypeScript/wiki/FAQ
Please fill in the *entire* template below.
-->
<!-- Please try to reproduce the issue with `typescript@next`. It may have already been fixed. -->
**TypeScript Version:** Version 3.6.3, Version 3.7.0-dev.20190911
<!-- Search terms you tried before logging this (so others can find this issue more easily) -->
**Search Terms:** intersect function
**Code**
```ts
class A {
public uniqueToA = "a";
public getString(this: A): string {
return "a";
}
}
class B {
public uniqueToB = "b";
public getString(this: B): string {
return "b";
}
}
function f(model: A | B): string {
return model.getString();
}
```
**Expected behavior:**
In Typescript 3.5.3, the `model.getString()` in `function f` has no error.
**Actual behavior:**
In Typescript 3.6.3 and 3.7.0-dev.20190911, the code above gets the following errors:
```
index.ts:18:12 - error TS2684: The 'this' context of type 'A | B' is not assignable to method's 'this' of type 'A & B'.
Type 'A' is not assignable to type 'A & B'.
Property 'uniqueToB' is missing in type 'A' but required in type 'B'.
18 return model.getString();
~~~~~
index.ts:10:12
10 public uniqueToB = "b";
~~~~~~~~~
'uniqueToB' is declared here.
```
**Playground Link:** <!-- A link to a TypeScript Playground "Share" link which demonstrates this behavior -->
http://www.typescriptlang.org/play/#code/MYGwhgzhAECC0G8BQ1XQA4FcBGICWw0mAdngI6YCmAKgPbwC80ARGMwNxIppa4HQBzSgBcAysIBOeYgIAUwgBZ4IALjgBKNREnSBibmjQSRmCcRZtOhgL5JbSUJBgAhfYd75CJclTqumzNgcXO44noIi4lIy8kqq0M6a0NrResiGhsbCpuaBwTZ2XABmJMDCeLTmRbIAtrQAJpQgavAAPglJKbpuRiZm0HWNIAB0QmI6Meqc1kA
**Related Issues:** #32506 <!-- Did you find other bugs that looked similar? -->
| Suggestion,Needs Proposal | low | Critical |
492,454,411 | flutter | RenderViewport.useTwoPaneSemantics is not show on all semantics nodes of a dropdown button | In the following code snippet, on the second activation of the dropdown, RenderViewport.useTwoPaneSemantics is only shown on one of the semantics nodes instead of all of them.
```dart
import 'package:flutter/material.dart';
void main() => runApp(MyApp());
class MyApp extends StatefulWidget {
@override
_MyAppState createState() => _MyAppState();
}
class _MyAppState extends State<MyApp> {
String value = 'foo';
@override
Widget build(BuildContext context) {
return MaterialApp(
title: 'Flutter Demo',
theme: ThemeData(
primarySwatch: Colors.blue,
),
home: Navigator(
onGenerateRoute: (_) {
return MaterialPageRoute(
builder: (context) {
return Scaffold(
body: SafeArea(
child: Center(
child: Row(
mainAxisAlignment: MainAxisAlignment.center,
mainAxisSize: MainAxisSize.min,
children: <Widget>[
DropdownButton(
value: value,
items: [
DropdownMenuItem(value: "foo", child: Text("Foo"),),
DropdownMenuItem(value: "bar", child: Text("Bar"),),
DropdownMenuItem(value: "baz", child: Text("Baz"),),
],
onChanged: (String value) { setState(() {
this.value = value;
});},
),
MaterialButton(child: Text('Reset'), onPressed: () {
setState(() {
value = 'foo';
});
}),
],
),
),
),
);
},
);
},
),
);
}
}
``` | framework,f: material design,a: accessibility,f: scrolling,P2,team-design,triaged-design | low | Minor |
492,474,934 | go | x/build/cmd/coordinator: merge double error returns into one error | Many parts of x/build return two errors, the "remote error" vs the "exec error". The first represents a command that executed remotely (but failed), and the latter is an error trying to even do the execution, and didn't even making to finding out the remote buildlet's exit status.
In the x/build these go by several names (rerr, err, network error, communication error) etc.
But the code gets ugly dealing with and propagating the two types of errors and I suspect not many points in the code care (retries, etc).
Now that Go 1.13 errors are a bit more mature (https://golang.org/doc/go1.13#error_wrapping) it's probably time to clean up x/bulid's errors and collapse all the double error returns back into one, where the "it was a failure to communicate" bit (or alternatively, perhaps easier: "the successfully failed" bit) are just a property on the error that we can ask about questions about when a caller cares.
/cc @dmitshur (who I can tell from his recent CLs has been hating this)
| Builders,NeedsInvestigation | low | Critical |
492,484,818 | flutter | Icon widget wrapped inside a Hero widget does not interpolate size or color(only position works) | The container which wraps the Icon widget works as expected but not the Icon. When I swap the Icon widget with a FlutterLogo widget, everything works. (tested on iOS emulator)
first.dart
```dart
import 'package:flutter/material.dart';
import 'package:transitiontest/second.dart';
class FirstPage extends StatefulWidget {
@override
_FirstPageState createState() => _FirstPageState();
}
class _FirstPageState extends State<FirstPage> {
@override
Widget build(BuildContext context) {
return Scaffold(
appBar: AppBar(
title: Text("First Page"),
),
body: Center(
child: ListView.builder(
itemCount: 100,
//itemExtent: 100.0,
itemBuilder: (context, index) {
return _StoryTile(index: index);
},
)),
);
}
}
class _StoryTile extends StatelessWidget {
final int index;
const _StoryTile({Key key, @required this.index}) : super(key: key);
@override
Widget build(BuildContext context) {
return InkWell(
onTap: () {
Navigator.of(context).push(
MaterialPageRoute(builder: (context) {
return SecondPage(index: index);
}),
);
},
child: Card(
margin: EdgeInsets.fromLTRB(16, 8, 16, 8),
color: Colors.purple,
elevation: 6.0,
shape:
RoundedRectangleBorder(borderRadius: BorderRadius.circular(16.0)),
child: Row(
mainAxisAlignment: MainAxisAlignment.start,
crossAxisAlignment: CrossAxisAlignment.center,
mainAxisSize: MainAxisSize.max,
children: <Widget>[
Padding(
padding: const EdgeInsets.all(10.0),
child: Hero(
transitionOnUserGestures: true,
tag: index,
child: Container(
//constraints: BoxConstraints.tightForFinite(),
color: Colors.blue,
child: Icon(
Icons.account_circle,
//key: Key("$index"),
color: Colors.white,
size: 40.0,
),
),
),
),
Text(
"Element: $index",
style: TextStyle(color: Colors.white, fontSize: 20.0),
)
],
),
),
);
}
}
```
second.dart
```
import 'package:flutter/material.dart';
class SecondPage extends StatefulWidget {
final int index;
const SecondPage({Key key, @required this.index}) : super(key: key);
@override
_SecondPageState createState() => _SecondPageState();
}
class _SecondPageState extends State<SecondPage> {
@override
Widget build(BuildContext context) {
return Scaffold(
body: Container(
constraints: BoxConstraints.expand(),
color: Colors.grey,
child: SafeArea(
child: Center(
child: Container(
//constraints: BoxConstraints.tight(Size(200.0, 100.0)),
//width: 200.0,
//height: 100.0,
child: Column(
mainAxisAlignment: MainAxisAlignment.center,
crossAxisAlignment: CrossAxisAlignment.center,
mainAxisSize: MainAxisSize.max,
children: <Widget>[
Text(
"Element ${widget.index} detail",
style: TextStyle(fontSize: 20.0, color: Colors.white),
),
SizedBox(height: 50.0),
Hero(
transitionOnUserGestures: true,
tag: 0,
child: Container(
//constraints: BoxConstraints.tightForFinite(),
color: Colors.blue,
child: Icon(Icons.account_circle,
size: 85.0, color: Colors.white),
),
),
SizedBox(height: 50.0),
Hero(
transitionOnUserGestures: true,
tag: 1,
child: Icon(Icons.cloud_circle,
size: 85.0, color: Colors.purple),
),
SizedBox(height: 50.0),
Hero(
transitionOnUserGestures: false,
tag: 2,
child:
Icon(Icons.cloud_circle, size: 85.0, color: Colors.pink),
),
],
)),
),
),
),
floatingActionButton: SafeArea(
child: FloatingActionButton(
onPressed: () {
Navigator.of(context).pop();
},
child: Icon(Icons.chevron_left),
),
),
);
}
}
```
| c: new feature,framework,a: animation,has reproducible steps,P3,found in release: 3.3,found in release: 3.6,team-framework,triaged-framework | low | Major |
492,489,747 | pytorch | sccache stats can cause whole build to fail | ## 🐛 Bug
A recent PR (#24506) added an incidental "sccache --show-stats" command in the pytorch
jenkins build script. Fine, but this command isn't universally present and this new stat
command can cause the whole build to unnecessarily fail, for example on a
recent ppc64le platform build.
## To Reproduce
Step to reproduce the behavior:
1. run the .jenkins/pytorch/build.sh script on ppc64le platform (as one example).
Resulting output snippet:
```
+ echo 'PyTorch Build Statistics'
PyTorch Build Statistics
+ sccache --show-stats
.jenkins/pytorch/build.sh: line 182: sccache: command not found
+ cleanup
```
## Expected behavior
Command to show sccache stats shouldn't have to force build failure. This was added
recently (last month). In other places within the build scripts, sccache is queried for
existence before use.
## Environment
Ubuntu, python 3.6, ppc64le
(However, python version is irrelevant here; failure related only to presence of sccache).
Other Environment info I believe to also be irrelevant.
I have a small patch for this (checking for presence of sccache); PR to follow shortly.
Injected as an incidental change as part of https://github.com/pytorch/pytorch/pull/24506 | module: build,triaged | low | Critical |
492,498,429 | flutter | [web]: 'flutter create --template=plugin' command should generate code for web | It would be great if code generated by 'flutter create --template=plugin' command contains a minimal sample for web. | c: new feature,tool,platform-web,P2,team-web,triaged-web | low | Minor |
492,517,053 | go | go/types: CheckExpr doesn't return partial results | <!-- Please answer these questions before submitting your issue. Thanks! -->
### What version of Go are you using (`go version`)?
<pre>
$ go version
go version go1.13 darwin/amd64
</pre>
### Does this issue reproduce with the latest release?
Yes
### What operating system and processor architecture are you using (`go env`)?
Darwin/amd64
### What did you do?
I used `types.CheckExpr` to check the expression "foo{i}" in the package:
```go
package fake
type foo struct {}
var _ = foo{i}
```
### What did you expect to see?
I expected to have some way to type check resiliently and get the type for my `*ast.CompositeLit`.
See https://play.golang.org/p/dDqvcc-jSFM
### What did you see instead?
The error for the un-type checkable "i" identifier is returned and no type checking is done. When using `*types.Checker` you can set the "Error" callback in the config so the type checker doesn't bail out, but CheckExpr doesn't set "Error" in the config.
I was experimenting with `types.CheckExpr` in gopls to provide completions in the face of certain kinds of syntax errors, but its utility is limited because it isn't resilient to errors. | NeedsInvestigation | low | Critical |
492,557,486 | opencv | function abortOnCannotGrowMemory() called on opencvjs after few minutes of run browser. | Hello,
This function `function abortOnCannotGrowMemory()` called in opencv.js after few minutes successful run of javascript package. Is there any way we can avoid "abort" so package can run smoothly. (I used TRY and CATCH but no use).
```
try{
cv.calcOpticalFlowPyrLK(oldGray, frameGray, p0, p1, st, err, winSize, maxLevel, criteria);
}catch(e){
console.log("abortOnCannotGrowMemory error coming",e)
}
```
Or, Is there an way we can increase or grow memory dynamically from programmatic level. So cleaning and reusing existing memory dynamically will help our package execution on browser.
| category: javascript (js) | low | Critical |
492,599,455 | flutter | Google Maps: smooth transition between coordinate points | I am creating a `GoogleMap` widget and trying to "animate" an icon on the map by continuously updating its position (lat, long). The icon is drawn in the center of the google map container using a `Stack` and it is the map that is moving underneath.
The problem is that I am not being able to achieve a smooth movement between position refresh.
**************** The code:
**- GoogleMap widget:**
```
Widget _buildGoogleMaps() {
return GoogleMap(
mapType: MapType.satellite,
initialCameraPosition: initialPos,
onMapCreated: (GoogleMapController controller) {
_mapController = controller;
});
}
```
**- Getting a new position:**
```
newPosition = CameraPosition(
target: LatLng(lat, lon),
zoom: 19,
);
_goToPosition(newPosition);
```
**- Moving to the new position:**
```
void _goToPosition(CameraPosition newPosition) {
// _mapController.animateCamera(
// CameraUpdate.newCameraPosition(newPosition),
// );
_mapController
?.moveCamera(CameraUpdate.newLatLng(LatLng(newPosition.target.latitude, newPosition.target.longitude)));
```
**************** The problems:
- when using `mapController.moveCamera` the map "jumps" from previous position, directly to the new one, so, when updating position like that multiple times, is looks like a movie in slow rate frame, moving jump by jump, not smooth.
- when using `mapController.animateCamera `method and its variants, the camera seems to move more smoothly but when a position out of the map container is selected, the map is blurry, seems to not be loaded when the camera reaches this position, so the icon appears navigating over a pixelated space.
Any suggestions/help/improvements to achieve this desired smooth-moving effect on the map?.
My approach might be totally wrong so I am open to any suggestion.
| d: examples,p: maps,package,team-ecosystem,P3,triaged-ecosystem | low | Major |
492,616,459 | pytorch | [Feature Request] Trace / Script C++ models | Hey,
does anyone know about a possibility of exporting and importing a learned model in C++?
I want to infer the net in a c++ project, where I don't have access to the class, which contains the net or the forward pass.
I found the solution to use:
```
string model_path = "model.pt";
serialize::OutputArchive output_archive;
model.save(output_archive);
output_archive.save_to(model_path);
```
But the outcome is a zip file with a .json header and just the weight and bias tensors as a file. When open with Netron, it just shows me the saved tensors without any connection. I tried to import the C++ exported model to Python to use there the ONNX export. But unfortunatly to that it needs the whole method of the forward pass, because it is not saved in the exported file. I would like to export a whole net in C++, including forward pass, netstructure ... like it is in the ONNX format.
Is there any solution to that kind of problem?
cc @suo @yf225 | oncall: jit,module: cpp,triaged | low | Major |
492,618,573 | ant-design | Tabs: pin one tab in header | - [ ] I have searched the [issues](https://github.com/ant-design/ant-design/issues) of this repository and believe that this is not a duplicate.
### What problem does this feature solve?
tabs组件支持固定其中某一项吗,当我有很多数据的时候,我希望其中某一项一直固定到最前面,而后面的能够滑动
### What does the proposed API look like?
通过传递tab的key进行控制是否固定某一项或者几项,固定位置不能够滑动
就像下面这个链接地址:[https://codesandbox.io/s/bgol8](https://codesandbox.io/s/bgol8)
我想把其中的tab0固定住
<!-- generated by ant-design-issue-helper. DO NOT REMOVE -->
| Inactive | low | Major |
492,644,561 | PowerToys | Cycle order of running applications (Alt-tab like ... but different) | I couldn't find it in the documentation, so I guess this is a feature request:
I would like to be able to switch focus between windows on a specific desktop, by cycling the focus of each window.
This is a little different from `ALT + TAB` because in `ALT + TAB` the focus isn't switched until you leave the keys, and repeated presses change the cycle order (the focus is switched back to the window you came from). I'm asking for something like 'focus next window in desktop' and 'focus previous window in desktop'. This is a common feature in tiling window managers in linux (for example, [AwesomeWM](https://github.com/awesomeWM/awesome)).
Does this feature exist in current PowerToys? | Idea-New PowerToy | low | Major |
492,653,500 | electron | Support authentication with session.resolveProxy() | <!-- As an open source project with a dedicated but small maintainer team, it can sometimes take a long time for issues to be addressed so please be patient and we will get back to you as soon as we can.
-->
### Preflight Checklist
<!-- Please ensure you've completed the following steps by replacing [ ] with [x]-->
* [x] I have read the [Contributing Guidelines](https://github.com/electron/electron/blob/master/CONTRIBUTING.md) for this project.
* [x] I agree to follow the [Code of Conduct](https://github.com/electron/electron/blob/master/CODE_OF_CONDUCT.md) that this project adheres to.
* [x] I have searched the issue tracker for a feature request that matches the one I want to file, without success.
### Problem Description
I'm using `session.resolveProxy()` to add proxy support to VS Code's extensions. That is currently missing support for authentication because `session.resolveProxy()` does not return credentials stored in the OS' proxy settings.
### Proposed Solution
`session.resolveProxy()` should include credentials stored in the OS' proxy settings.
### Alternatives Considered
Someone should write a Node module doing this. Getting it right across platforms might be the main challenge. Not sure if Chromium's implementation for reading proxy settings could be extracted to make it reusable as a Node module.
| enhancement :sparkles: | low | Minor |
492,670,535 | terminal | Compatibility Issue with Holdkey | # Environment
```none
Windows build number: Microsoft Windows [Version 10.0.18362.356]
Windows Terminal version (if applicable): N/A
Any other software? Holdkey (http://www.holdkey.eu/) and WSL Ubuntu 18.04 LTS
```
# Steps to reproduce
Install `holdkey` from [holdkey.eu](http://www.holdkey.eu/). Open a WSL terminal and try to type a path.
# Expected behavior
Typing the `/` key will yield a `/` character.
# Actual behavior
When `holdkey` is running, typing the `/` key on a US keyboard layout yields a `&` character.
When `holdkey` is not running, the `/` character is yielded.
# Other information
I've reported this bug to the `holdkey` team. This behavior is only in the WSL Terminal, other applications work fine.
# EDIT
This also happens in the `command prompt` but not in `powershell` | Product-Conhost,Area-Input,Issue-Bug,Priority-2 | low | Critical |
492,674,808 | TypeScript | Allow hooks for watch events | ## Search Terms
- hooks
- events
- watch
## Suggestion
Allow a ts file to react to actions performed by tsc while it is running in ```--watch``` mode. This can be done by having that file export specific members of specific known types that tsc will execute at the appropriate time.
## Use Cases
I would like to automatically regenerate an index.ts file before letting tsc proceed with its normal compilation that would include the now altered index.ts.
In one of my projects, I would also want to restart a server after compilation. Others may want to do other build related tasks, such as re-running a subset of tests after build for example.
## Examples
Given a command line
```tsc --build --watch --watchHook ./watcher.ts```
and the following watcher.ts file:
```typescript
/*
"tsc-watch" is a tentative name for a lib that would define the "TS*"
types used below.
*/
/// <reference lib="tsc-watch" />
interface TSChangeSet {
[key: string]: {
files: string[];
buildInfo: TSBuildInfo;
};
}
const hooks: TSWatcherHooks = {
/**
* Executes before an initial build.
*
* This should also be executed once for each referenced project.
*
* @param tsconfigPath An absolute path to the tsconfig being built.
* This is most useful to distinguish between projects in references.
* @param buildInfo undefined if there was no prior buildInfo file.
* Else, it's an object representation of its contents before compilation.
*
* @return If the returned value is a promise, it will be awaited before
* proceeding. If it is anything else, tsc will resume immediately.
* If the promise is rejected or this handler throws, the compilation
* should be considered to have failed.
*/
onBeforeBuild: (
tsconfigPath: string,
buildInfo?: TSBuildInfo
): unknown | Promise<unknown> => {
return;
},
/**
* Executes after an initial build.
*
* This should also be executed once for each referenced project.
*
* @param tsconfigPath An absolute path to the tsconfig being built.
* This is most useful to distinguish between projects in references.
* @param buildInfo undefined if not an incremental build.
* Else, it's an object representation of its contents after compilation.
*
* @return If the returned value is a promise, it will be awaited before
* proceeding. If it is anything else, tsc will resume immediately.
* If the promise is rejected or this handler throws, tsc should exit
* with an error code.
*/
onAfterBuild: (
tsconfigPath: string,
buildInfo?: TSBuildInfo
): unknown | Promise<unknown> => {
return;
},
/**
* Executes before a rebuild.
*
* This should be executed once regardless of whether the change is in just
* one project or multiple.
*
* If there are additional changes after this handler (e.g. a file is
* edited as a result of this call), then compilation should NOT proceed,
* and yet this handler should be executed a second time with the newly
* changed files.
*
* @param changedFiles A set of objects changed since last build.
* Each key in the set is a path to a tsconfig file.
* Each value is an object with the build info as one member,
* and an array of changed files as the other.
*
* @return If the returned value is a promise, it will be awaited before
* proceeding. If it is anything else, tsc will resume immediately.
* If the promise is rejected or this handler throws, the recompilation
* should be considered to have failed.
*/
onBeforeRebuild: (changedFiles: TSChangeSet): unknown | Promise<unknown> => {
// Do stuff before recompilation, e.g. (re)generate index files
return;
},
/**
* Executes after a rebuild.
*
* This should be executed once regardless of whether the change is in just
* one project or multiple.
*
* @param changedFiles A set of objects changed since last build.
* Each key in the set is a path to a tsconfig file.
* Each value is an object with the build info as one member,
* and an array of changed files as the other.
*
* All changes from before handlers are also merged into this set.
*
* @return If the returned value is a promise, it will be awaited before
* proceeding. If it is anything else, tsc will resume immediately.
* If the promise is rejected or this handler throws, tsc should exit
* with an error code.
*/
onAfterRebuild: (changedFiles: TSChangeSet): unknown | Promise<unknown> => {
// Do stuff after recompilation, e.g. run test subset, restart app, etc.
return;
}
};
export default hooks;
```
I would like tsc to behave as described in the watcher.ts file. As with most TS settings, there should be an analog in the tsconfig.json file, where the path to the hooks file is relative to the tsconfig.json file.
Currently, in order to do these sorts of things, one has to use a separate builder like gulp, which is somewhat complex to use for projects that use project references.
## 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 |
492,724,484 | PowerToys | Display system tray (icons and language bar) on multiple monitors | Currently, tray icons (e.g. anti-virus and network icons), as well as the language bar icon, appear only on the "main display".
If I need to check the status of one of these icons (e.g. what's the current typing language? Is my speaker muted?), or click one of them (e.g. open OneNote by clicking its tray icon), and I'm looking at the "wrong" screen, I have to move my attention and my mouse to the other screen.
I know there are third-party utilities that do this, but I always thought it was supposed to be Windows' responsibility.
systray | Idea-New PowerToy,Product-Tweak UI Design | medium | Major |
492,771,761 | rust | #[global_allocator] in dependent crate can be optimized away | `jemallocator-global` is a library intended to make jemalloc your allocator. It works by declaring a static allocator with `#[global_allocator]`.
However, a downstream crate linked with `jemallocator-global` nonetheless uses the system malloc on MacOS X.
After experimenting a bit with `#[global_allocator]`, it's behaving as though rustc optimizes away a whole crate if you don't seem to be using it; and it doesn't detect that the global allocator is always used. | A-linkage,A-allocators,T-compiler,C-bug | low | Major |
492,773,143 | terminal | Content sometimes disappears when moving back to Terminal window | <!--
🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨
I ACKNOWLEDGE THE FOLLOWING BEFORE PROCEEDING:
1. If I delete this entire template and go my own path, the core team may close my issue without further explanation or engagement.
2. If I list multiple bugs/concerns in this one issue, the core team may close my issue without further explanation or engagement.
3. If I write an issue that has many duplicates, the core team may close my issue without further explanation or engagement (and without necessarily spending time to find the exact duplicate ID number).
4. If I leave the title incomplete when filing the issue, the core team may close my issue without further explanation or engagement.
5. If I file something completely blank in the body, the core team may close my issue without further explanation or engagement.
All good? Then proceed!
-->
<!--
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
Windows build number: 10.0.18362.0
Windows Terminal version (if applicable): 0.4.2382.0
Any other software?
```
# Steps to reproduce
Can't reproduce consistently, but sometimes, when I navigate back to Windows Terminal from another window, the content of the Terminal window completely disappears and the window becomes transparent.
I'm using PowerShell Core, and I typically find that the problem occurs after running an `npm` script. Sometimes, if I hit up and then Enter, the previous command re-runs and the window returns to normal. On other occasions, this seems to have no effect.
| Help Wanted,Needs-Repro,Issue-Bug,Area-UserInterface,Product-Terminal,Priority-3 | low | Critical |
492,801,110 | neovim | Asynchronous `:write` | When working with `sshfs` or other remote filesystems, sometimes writes take many seconds to complete. It would be nice if the write could be performed asynchronously so the UI remained responsive while that was happening - perhaps because the user wishes to continue work editing another buffer or whatever. On the UI they'd know the write had completed because the `modified` state will disappear.
Taking inspiration from `:&grep` maybe this could be spelled `:&write` or `:&w`.
Optionally, a setting to make async writes default on `:w` would be handy too, maybe spelled `:set asyncwrite` | enhancement,io | low | Major |
492,870,026 | pytorch | [RFC] TensorBoard extensions and improvements for PyTorch | ## 🚀 TensorBoard Improvements
In collaboration with Google, we've made great progress on adding support for PyTorch in TensorBoard. If you've been using `torch.utils.tensorboard` and the TensorBoard server with PyTorch training that is excellent! Hopefully you noticed that we support a wide range of visualization options from scalars and histograms to 3D meshes.
Now that we've added baseline support, we are curious how things can be improved.
We recently opened a few issues on TensorBoard's GitHub repo:
* **C++ logging from non-TensorFlow frameworks** - https://github.com/tensorflow/tensorboard/issues/2636
* **Insights embedded in TensorBoard** - https://github.com/tensorflow/tensorboard/issues/2637
* **Loading indicator for slow progress** - https://github.com/tensorflow/tensorboard/issues/2638
We are also exploring:
* **Training-time metrics (CPU, GPU, memory, etc) in TensorBoard**
If you have other ideas or feel any of the above would benefit you, we'd love to hear from you!
cc @nfelt @manivaradarajan @GalOshri @natalialunova @lanpa @sanekmelnikov @caraya10 @jspisak @soumith
| triaged,module: tensorboard | low | Major |
492,883,324 | go | cmd/go: do not modify the 'go.mod' or 'go.sum' files if the 'go' directive indicates a newer-than-supported version? | @jayconrod notes in https://github.com/golang/go/issues/34217#issuecomment-530873990:
> One problem with this: adding `go 1.14` to go.mod does not prevent old versions of the Go command from adding `+incompatible` versions. So a module could be broken unintentionally in a mixed environment.
There is a more general version of this problem: older versions of the Go toolchain may not know how to correctly interpret a `go.mod` file intended for a newer version, and thus could make arbitrarily wrong edits.
Probably we should have the `go` command error out instead of making changes if it thinks a `go.mod` file generated by a newer toolchain is incomplete or inaccurate. | NeedsFix,GoCommand,modules | low | Critical |
492,903,789 | flutter | After upgrading to 1.9 `flutter logs` on iOS displays system logs | After upgrading to Flutter 1.9, `flutter logs` on a connected iOS device shows the full system logs, and not the filtered logs for Flutter as it was with previous versions.
```
Doctor summary (to see all details, run flutter doctor -v):
[✓] Flutter (Channel beta, v1.9.1+hotfix.2, on Mac OS X 10.14.5 18F203, locale en-PT)
[✓] Android toolchain - develop for Android devices (Android SDK version 29.0.0)
[✓] Xcode - develop for iOS and macOS (Xcode 10.2.1)
[✓] Android Studio (version 3.4)
[✓] VS Code (version 1.38.1)
[✓] Connected device (1 available)
• No issues found!
```
The same occurs in stable | platform-ios,tool,P3,team-ios,triaged-ios | low | Major |
492,918,250 | rust | Document how ptr::drop_in_place behaves when dropping a slice panics | When dropping a slice `[T]` using `ptr::drop_in_place`, a `T::drop` can panic. I can't find any documentation of what the behavior of the program is in that case.
| C-enhancement,A-FFI,T-lang,A-docs | low | Minor |
492,925,058 | go | net: spurious EADDRINUSE from connect syscall on FreeBSD when connecting from IPv6 wildcard to IPv4 address | Observed in a `freebsd-amd64-12_0` TryBot: https://storage.googleapis.com/go-build-log/2e4c3b9c/freebsd-amd64-12_0_d1b5be43.log
```
--- FAIL: TestDialerLocalAddr (0.00s)
dial_test.go:647: tcp [::]:0->127.0.0.1: got dial tcp [::]:0->127.0.0.1:22464: connect: address already in use; want <nil>
FAIL
FAIL net 54.481s
```
This test is also flaky on macOS (#22019), but the symptom on FreeBSD is different from the one observed on macOS (`connect: address already in use` vs. `getsockopt: operation timed out`).
Note that the `freebsd-arm-paulzhol` flake reported in https://github.com/golang/go/issues/22019#issuecomment-355091973 matches this one.
CC @ianlancetaylor @mikioh | ExpertNeeded,OS-FreeBSD,NeedsInvestigation | low | Major |
492,942,091 | terminal | Setting cursor position has too many routes | Based on #2731, repeated here:
>>When it came to actually setting the cursor position, I found there were a number of options to choose from, with quite different behaviours. Some operations use a variation of SetConsoleCursorPosition, some use AdjustCursorPosition, and some call the SCREEN_INFORMATION::SetCursorPosition method directly, or even just Cursor::SetPosition. It wasn't always clear to why a particular method was chosen.
>
>This certainly needs a follow up work item. I'm reviewing some of the code in each of these to try to come up with reasoning why you should use one over the other... and I'm realizing that this is probably the source of our "the cursor is blinking weirdly or is on when it shouldn't be while text is outputting" problems.
>
>As far as I can tell, you've chosen the correct one with SCREEN_INFORMATION::SetCursorPosition because that one is responsible for updating the delay status on the cursor so it stays off or on instead of blinking wildly as it is moving rapidly around the screen when things are happening. The cursor is supposed to temporarily suspend blinking when there is activity (by the delay).
>
>I am certain the inconsistency in application of which cursor-setting-function is being used is responsible for graphical glitches.
This work item represents auditing the cursor position setting methods to reconcile them into a (hopefully) more unified entry point. It probably also represents analyzing the assorted delay booleans that temporarily suspend cursor drawing to reconcile those as well. (`Cursor::_fDelay` and `Cursor::_fIsVisible` and `Cursor::_fBlinkingAllowed` and `Cursor::fIsOn`. and `Cursor::f_DeferCursorRedraw`.) | Product-Conhost,Help Wanted,Area-Interaction,Issue-Bug | low | Major |
492,957,872 | kubernetes | Reconsider using etcd+chunking for reflector lists | Filing issue for the discussion I brought up in SIG scalability meeting today, as I couldn't find one for it already.
**Background**:
Currently all our informers, which use reflectors underneath, list from apiserver's watch cache. This was a decision made quite some time ago to use watch cache instead of etcd - to avoid overloading etcd with several lists.
Now, since the watch cache doesn't support chunking (due to lack of history), informers make a full list (without chunking) to the apiserver. This is causing an issue that we recently encountered where multiple watches trying to list at the same time caused apiserver CPU/memory to spike quite a bit:
And indeed such a scenario can happen if for e.g the watch from apiserver to etcd doesn't receive an event before that resource version gets compacted from etcd:
```
E0824 09:43:50.201103 8 watcher.go:208] watch chan error: etcdserver: mvcc: required revision has been compacted
W0824 09:43:50.201169 8 reflector.go:256] storage/cacher.go:/pods: watch of *core.Pod ended with: The resourceVersion for the provided watch is too old.
W0824 09:43:51.201450 8 cacher.go:125] Terminating all watchers from cacher *core.Pod
```
And as a result when apiserver terminates all watches from the cacher, that caused multiple clients to relist at the same time, causing the resource spike on apiserver.
From our discussion, it seems that apiserver watch falling out of etcd window can happen in at least two scenarios currently:
- set of objects being watched are not changing frequently enough, causing that latest RV to go out of etcd's window. @wojtek-t - IIUC this issue will be solved with an etcd 3.4's feature you mentioned that's similar to our watch bookmark. Can you confirm?
- watch cache not able to handle the throughput of incoming events from etcd. This problem FMU still remains even after moving to etcd 3.4. So to avoid apiserver spike in such cases, IMO we should reconsider the option of allowing informers to list from etcd with chunking (since we have it now) instead of watch cache. It seems feasible, per Wojtek's thoughts, to be able to do this starting from etcd 3.4 as it has significant read concurrency improvements
**What would you like to be changed**:
- Switch to etcd 3.4
- Allow informers to list from etcd (with chunking)
- Scale test it
cc @kubernetes/sig-scalability-feature-requests @wojtek-t @liggitt @smarterclayton | sig/scalability,kind/feature,lifecycle/frozen | low | Critical |
492,959,466 | pytorch | Support the AVX512 runtime dispatch | ## 🚀 Feature
<!-- A clear and concise description of the feature proposal -->
Enhance the runtime dispatch to support AVX512
## Motivation
<!-- Please outline the motivation for the proposal. Is your feature request related to a problem? e.g., I'm always frustrated when [...]. If this is related to another GitHub issue, please link here too -->
We are optimizing some quantized kernels by using AVX2 or AVX512 instruction set. It gave significant performance win for server side quantized model. Currently in OSS, we already have support for dispatching for AVX and AVX2 in pytorch/aten/src/ATen/native/DispatchStub.h.
## Pitch
<!-- A clear and concise description of what you want to happen. -->
## Alternatives
<!-- A clear and concise description of any alternative solutions or features you've considered, if any. -->
## Additional context
<!-- Add any other context or screenshots about the feature request here. -->
cc @jerryzh168 @jianyuh @dzhulgakov @raghuramank100 @jamesr66a | feature,low priority,triaged | low | Major |
492,959,591 | bitcoin | Syncing headers with feeler-peers | It occurred to me that we should eliminate the distinction between feelers -- peers we connect to on average every couple minutes to test entries in addrman -- and the extra outbound peers we connect to if our tip is stale (#11560). Instead, why not just sync headers with feeler connections, and if we happen to learn of a new block from that feeler when our tip had been stale, consider evicting an older connection (just as we do with our extra outbounds)?
It seems to me that this should increase the cost to an adversary trying to eclipse a node, because they'd need to control a large fraction of the addrman in order to withhold the most-work chain for any sizeable amount of time.
Moreover since we're already making these feeler connections already, it should be a small amount of additional bandwidth to sync headers with such a peer. (Perhaps we should avoid doing this in IBD...?)
@EthanHeilman Any thoughts? | Brainstorming,P2P | low | Major |
492,962,215 | pytorch | Access data_ptr in RNN.cpp | https://github.com/pytorch/pytorch/blob/master/aten/src/ATen/native/RNN.cpp#L598 is trying to access `data_ptr` of Tensor inside `TypeDefault` context. This won't work for backends like XLA. In general if we want to access `data_ptr`, it's preferred to be a backend implementation.
In fact in this case it seems `batch_sizes` can be a `int[]` instead of `Tensor`. But we need to consider BC for this change.
Related issue: https://github.com/pytorch/xla/issues/927
there's a PR but it's not a preferred way to fix: https://github.com/pytorch/pytorch/pull/26056
cc: @gchanan
cc @zou3519 | module: rnn,triaged | low | Major |
492,969,760 | flutter | Reference to "old widget" when navigating with PageRouteBuilder (or others) | ### Feature ###
The idea is to have `Navigator` save a reference or a copy of the current base widget when pushing a new one. Kinda like a Hero Widget. The added flexibility for transitions would make a lot of things possible!
PageRouteBuilder makes the "new widget" enter screen with an animation, but the "old widget" doesn't change at all.
Something like this is happening now:
> Old Page ---> Transitions (new widget superimpose old one) ----> new Page
Currently you can make it slide using `CupertinoPageRoute`, but IMHO a builder should grant some more flexibility.
I have successfully built this desired result once by passing my "old page" widget as an argument in my RouteSettings.
```
return PageRouteBuilder(
transitionDuration: const Duration(milliseconds: 1000),
pageBuilder: (context, animation, _) => widget,
transitionsBuilder: (context, animation, _, child) =>
leavingWidget == null ?
ScaleTransition(
scale: animation.drive(CurveTween(curve: Curves.decelerate)),
child: child,
) :
Stack(
children: [
SlideTransition(
position: Tween(
begin: Offset.zero,
end: Offset(-1.0, 0.0),
).animate(animation),
child: leavingWidget.widget,
),
SlideTransition(
position: Tween(
begin: Offset(1.0, 0.0),
end: Offset.zero,
).animate(animation),
child: widget,
),
],
),
// if (?:) statement
// transitionBuilder
);
```
So the idea is to have some optimization on getting the `leavingWidget` (in this example I used `Navigator.pushNamed( _, arguments: this)` in a widget's state.
The above transition makes the "old widget" slide out of the screen as the new one enters it. It removes the feeling of "overlapping" for a feeling of "pushing away".
### Conclusion ###
Request to have Navigator hold a reference or copy to the old page widget when pushing a new one for more animation options. | c: new feature,framework,f: routes,P3,team-framework,triaged-framework | low | Minor |
492,987,346 | angular | Adding a migration to replace TestBed.get with TestBed.inject | <!--🔅🔅🔅🔅🔅🔅🔅🔅🔅🔅🔅🔅🔅🔅🔅🔅🔅🔅🔅🔅🔅🔅🔅🔅🔅🔅🔅🔅🔅🔅🔅
Oh hi there! 😄
To expedite issue processing please search open and closed issues before submitting a new one.
Existing issues often contain information about workarounds, resolution, or progress updates.
🔅🔅🔅🔅🔅🔅🔅🔅🔅🔅🔅🔅🔅🔅🔅🔅🔅🔅🔅🔅🔅🔅🔅🔅🔅🔅🔅🔅🔅🔅🔅🔅🔅-->
# 🚀 Feature request
### Command (mark with an `x`)
<!-- Can you pin-point the command or commands that are relevant for this feature request? -->
<!-- ✍️edit: -->
```
- [ ] new
- [ ] build
- [ ] serve
- [ ] test
- [ ] e2e
- [ ] generate
- [ ] add
- [x] update
- [ ] lint
- [ ] xi18n
- [ ] run
- [ ] config
- [ ] help
- [ ] version
- [ ] doc
```
### Description
With the deprecation of `TestBed.get` (https://github.com/angular/angular/pull/32406), would it be a good addition to add a schematic to update the usages to `TestBed.inject`?
### Describe the solution you'd like
Something like the following?
https://github.com/angular/angular-cli/compare/master...timdeschryver:pr/testbed-migration?expand=1
### Describe alternatives you've considered
/
| area: testing,freq2: medium,P5 | low | Major |
492,991,783 | kubernetes | Migrate kubelet to use v1beta1 Events | Now that the scheduler is migrated to new Events API, we can migrate the kubelet to emit events against the new API
This is part of #76674
List of the files that needs to be migrated:
```
./pkg/kubelet/kubelet.go
./pkg/kubelet/prober/prober.go
./pkg/kubelet/kubelet_pods.go
./pkg/kubelet/config/config.go
./pkg/kubelet/images/image_gc_manager.go
./pkg/kubelet/pod_workers.go
./pkg/kubelet/network/dns/dns.go
./pkg/kubelet/kuberuntime/kuberuntime_manager.go
./pkg/kubelet/active_deadline.go
./pkg/kubelet/oom/oom_watcher_linux.go
./pkg/kubelet/container/helpers.go
./pkg/kubelet/preemption/preemption.go
./pkg/kubelet/kubelet_node_status.go
./pkg/kubelet/cm/node_container_manager_linux.go
./pkg/kubelet/eviction/eviction_manager.go
./pkg/kubelet/images/image_manager.go
```
/assign @tedyu @yastij @wojtek-t
FYI @dims
/sig instrumentation
/sig scalability
/sig node
/priority important-soon
| sig/scalability,sig/node,kind/feature,sig/instrumentation,do-not-merge/hold,lifecycle/frozen | low | Major |
492,995,128 | flutter | Support Google AdSense in Flutter Web applications | I tried to implement the next script(Google Ad-sense) at my Flutter Web site, but unsuccessful by now. The code is:
```
<script type="text/javascript">var nend_params={"media":00000,"site":000000,"spot":000000,"type":1,"oriented":1};0</script>
<script type="text/javascript" src="https://js1.nend.net/js/nendAdLoader.js"></script>
```
I understood, that Flutter web is currently in technical preview, but may be any route? Any advice and suggestions will be greatly appreciated. | c: new feature,would be a good package,customer: crowd,platform-web,package,c: proposal,P1,team-web,triaged-web | medium | Critical |
493,002,069 | flutter | There should be a test for interacting with an app after hot restart | This path is a critical user journey, a test for it would have caught #40303. | a: tests,team,framework,P3,team-framework,triaged-framework | low | Minor |
493,008,480 | pytorch | Tracing non-constant shapes is broken | ## 🐛 Bug
```
import torch
def foo(x):
return torch.full(x.shape, 2)
x = torch.randn(3)
ge = torch.jit.trace(foo, x)
y = torch.randn(2, 5)
z = torch.randn([])
>>> ge(x).shape # expected: 3
3
>>> ge(y).shape # expected: (2, 5)
2
```
Tracing non-constant shapes seem to be broken if the shapes have different dimensions. I ran the above code on both linux and windows on master.
cc @suo @eellison | oncall: jit,triaged | low | Critical |
493,011,462 | pytorch | Make GloballyUniqueId a common type for both rpc and dist autograd | Both RPC and Distributed Autograd require globally unique id in various places. #25499 partially implements that in `GloballyUniqueId` for RRefs, which can be shared between rpc and dist autograd after adding overflow checks and `int()` operator.
cc @pietern @mrshenli @pritamdamania87 @zhaojuanmao @satgera @gqchen @aazzolini @rohan-varma @xush6528 | triaged,better-engineering,module: rpc | low | Minor |
493,013,862 | pytorch | Process fails with assertion error in magma-cuda100 | ## 🐛 Bug
During training the process dies with the following error:
```
...
global_step=101292, batch=73, batch_group=0: 30%|███████████████████▏ | 74/250 [05:01<10:25, 3.56s/it]
python: /opt/conda/conda-bld/magma-cuda100_1564975479425/work/interface_cuda/interface.cpp:897: void magma_queue_create_from_cuda_internal(magma_device_t, cudaStream_t, cublasHandle_t, cusparseHandle_t, magma_queue**, const char*, const char*, int): Assertion `queue->dAarray__ != __null' failed.
[1] 71261 abort python train_reconstruct.py --resume
```
## To Reproduce
Steps to reproduce the behavior:
1. Training a fairly large network while almost using all GPU memory (32GB on a Tesla V100).
1. The dataloader is rendering using OpenGL and then returning CPU tensors.
<!-- If you have a code sample, error messages, stack traces, please provide it here as well -->
## Expected behavior
Program should not crash or if out of memory should return the relevant error.
## Environment
PyTorch version: 1.2.0
Is debug build: No
CUDA used to build PyTorch: 10.0.130
OS: Ubuntu 16.04.5 LTS
GCC version: (Ubuntu 5.4.0-6ubuntu1~16.04.11) 5.4.0 20160609
CMake version: version 3.5.1
Python version: 3.7
Is CUDA available: Yes
CUDA runtime version: Could not collect
GPU models and configuration:
GPU 0: Tesla V100-SXM2-32GB
GPU 1: Tesla V100-SXM2-32GB
GPU 2: Tesla V100-SXM2-32GB
GPU 3: Tesla V100-SXM2-32GB
GPU 4: Tesla V100-SXM2-32GB
GPU 5: Tesla V100-SXM2-32GB
GPU 6: Tesla V100-SXM2-32GB
GPU 7: Tesla V100-SXM2-32GB
Nvidia driver version: 410.48
cuDNN version: Could not collect
Versions of relevant libraries:
[pip3] numpy==1.17.1
[pip3] torch==1.2.0
[pip3] torchfile==0.1.0
[pip3] torchnet==0.0.4
[pip3] torchvision==0.4.0
cc @vishwakftw @SsnL @jianyuh | module: dependency bug,needs reproduction,module: crash,triaged,module: linear algebra | medium | Critical |
493,017,543 | godot | Bug in display of 64-bit integers in the debugger | **Godot version:**
Godot_v3.1.1-stable_win64
**Issue description:**
I saw some suspicious values when using the debugger with my project, so I wrote the following snippet to investigate it (putting a breakpoint on the "pass" to look at the debugger):
```
var r := RandomNumberGenerator.new()
var seeds := []
var nums := []
r.randomize()
for i in 5:
nums.append(r.randi())
seeds.append(r.get_seed())
print(seeds)
pass
```
The code works fine and printed (in one run):
[5283336804004298712, -9206889307139333561, 918266396063285450, -3754841591890928943, -1916577486000920820]
However, in the same run, when I looked at the seeds array in the debugger, what I saw was:
The numbers in the debugger are obviously close, but no cigar.
I reran this code several times and this bug is consistent. The least-significant 2-4 digits are always wrong. The last digit is always zero and the second-to-last digit is usually zero (but not always).
| enhancement,topic:editor,confirmed | low | Critical |
493,022,610 | vscode | "editor font zoom" does not persist | - VSCode Version:1.38
- OS Version:Win 10
"Editor font zoom in" and "out" does not persist. This is inconsistent with "view zoom" which does persist. I understand that this is basically a shortcut to the "editor.fontSize" setting, but in that same vein, view zoom is just a shortcut to change the "window.zoom level" setting
Duplicate
1. "editor: font zoom out"
2. close & restart
3. font will have reverted pre-zoom size
Edit: updated title and text to reflect correct command "editor font zoom in/out" | feature-request,accessibility,editor-core | low | Major |
493,028,714 | rust | io::Stderr should be line-buffered by default, not unbuffered | `std::io::stderr` does no buffering at all by default, which means that a Rust program that uses the most obvious way to print error messages (`eprintln!`) will make individual `write` syscalls not only for each call to the macro, but for each subcomponent of the formatted string:
```plain
$ cat test.rs
fn main() {
eprint!("partial ");
eprintln!("line");
eprintln!("a one and a two and a {} and a {}", 3, 4.0);
}
$ rustc test.rs
$ strace -e trace=write sh -c 'exec ./test 2> /dev/null'
write(2, "partial ", 8) = 8
write(2, "line\n", 5) = 5
write(2, "a one and a two and a ", 22) = 22
write(2, "3", 1) = 1
write(2, " and a ", 7) = 7
write(2, "4", 1) = 1
write(2, "\n", 1) = 1
```
This behavior is undesirable in any context where multiple programs might be emitting error messages to the same terminal, logfile, or whatever at the same time (for instance, `make -j`) because partial lines from different programs can get mixed up with each other. If stderr instead buffered up a full line and wrote it to the OS all at once, then different programs' output could only get mixed together on a line-by-line basis, which is usually much less confusing for a person reading the logs.
This behavior is also troublesome for programs that incrementally parse stderr output; for instance, it may be the reason why emacs' compilation-mode occasionally doesn't detect all of the diagnostics in `cargo build` output (I don't know if there's an existing bug report for this).
(There is a strong case for having `eprint!` flush out the partial line that it generates, but that could be handled inside of `eprint!`.)
(This is closely related to, but not the same as, #60673, which is about stdout. Block buffering is _not_ normally appropriate for `stderr`, even when writing to a file.) | T-libs-api,C-feature-request | low | Critical |
493,054,311 | electron | Implement setAlwaysOnBottom |
### Preflight Checklist
* [x] I have read the [Contributing Guidelines](https://github.com/electron/electron/blob/master/CONTRIBUTING.md) for this project.
* [x] I agree to follow the [Code of Conduct](https://github.com/electron/electron/blob/master/CODE_OF_CONDUCT.md) that this project adheres to.
* [x] I have searched the issue tracker for a feature request that matches the one I want to file, without success.
### Problem Description
I have a need to place a BrowserWindow to the very bottom, as it's supposed to be a desktop decoration. `type: 'desktop'` unfortunately won't work for me since I still want to be able to receive events on it, and it seems to not be supported on Windows.
### Proposed Solution
Implementing the `setAlwaysOnBottom` API that would do the opposite of the currently implemented `setAlwaysOnTop`.
### Alternatives Considered
N/A
### Additional Information
N/A
Thank you very much for maintaining this beautiful piece of software! | enhancement :sparkles:,component/BrowserWindow | low | Minor |
493,060,565 | pytorch | There should be gating around BFloat16 | As we add support for BFloat16 we should add some gating mechanism which will allow user to opt in to bfloat16 usage. The reason for this is that it will be a while till the devices which support bf16 will be mass available and before that will happen, the performance of bf16 operations wont be the best. | module: performance,triaged,enhancement,module: bfloat16 | low | Major |
493,065,257 | flutter | [web] allow controlling the window/iframe/screenshot parameters in engine tests | Screenshot tests in the engine use headless Chrome's default window size and pixel density, hard-code 500x500 iframe size, and always take the screenshot of the entire window. This makes tests less predictable. Let's allow tests customize this.
Let's try to capture as many of the following as we can:
- [ ] control window size
- [ ] control iframe size
- [x] control screenshot rect
- [ ] control screen pixel density
- [x] use the most predictable rasterizer available (although this might be the default already: https://bugs.chromium.org/p/chromium/issues/detail?id=765284)
/cc @ditman filing this issue so we can track it as part of the infra project. Feel free to adjust the description.
| a: tests,team,engine,platform-web,P2,team-web,triaged-web | low | Critical |
493,068,434 | pytorch | c10 List API hard to use | There isn't any way to access the values of Generic List except as a reference. When writing ops this makes you do:
```
auto val = pop(stack);
auto list = val.toGenericListRef();
```
instead of:
```
auto list = pop(stack).toGenericList();
```
If you try to do the latter:
```
auto input_list = pop(stack).toGenericList();
for (const auto& input : input_list) {
auto tup = input.toTuple()
}
```
then you get the following error:
> ‘const class c10::impl::ListElementReference<c10::IValue, __gnu_cxx::__normal_iterator<c10::IValue*, std::vector<c10::IValue, std::allocator<c10::IValue> > >, c10::IValue>’ has no member named ‘toTuple’
This makes it difficult to use GenericList as a value instead of a reference. | module: internals,triaged | low | Critical |
493,069,199 | godot | Translation keys on control nodes are very inconsistent | **Godot version:**
3.1.1 stable
**OS/device including version:**
Linux Mint 19.2
**Issue description:**
Putting translation keys on Controls is very inconsistent. Normally, nodes are supposed to shrink to fit the text, but all sorts of things break that and cause it to be large enough to fit the full translation key.
For example, the key `ui_options_quit_without_saving` has the text `Quit Without Saving`, which is obviously shorter. Immediately after creation it shrinks properly, but doing various different things seem to make it keep the full size of the key.
**Steps to reproduce:**
Open the reproduction project. Run the game; the top-left button will fit the text, but the bottom-right won't. Moving the top-left button at all will make it stop working properly too.
**Minimal reproduction project:**
[GodotIssue.zip](https://github.com/godotengine/godot/files/3607742/GodotIssue.zip) | bug,confirmed,topic:gui | low | Minor |
493,091,352 | TypeScript | navto does not include jsdoc @typedef symbols | <!-- 🚨 STOP 🚨 𝗦𝗧𝗢𝗣 🚨 𝑺𝑻𝑶𝑷 🚨
Half of all issues filed here are duplicates, answered in the FAQ, or not appropriate for the bug tracker. Even if you think you've found a *bug*, please read the FAQ first, especially the Common "Bugs" That Aren't Bugs section!
Please help us by doing the following steps before logging an issue:
* Search: https://github.com/Microsoft/TypeScript/search?type=Issues
* Read the FAQ: https://github.com/Microsoft/TypeScript/wiki/FAQ
Please fill in the *entire* template below.
-->
<!-- Please try to reproduce the issue with `typescript@next`. It may have already been fixed. -->
**TypeScript Version:** 3.7.0-dev.20190911
<!-- Search terms you tried before logging this (so others can find this issue more easily) -->
**Search Terms:**
- navTo
- workspace symbol search
**Code**
For the js file:
```ts
/**
* Describes a duck
* @typedef {{rubber: boolean, color: string}} Duck
*/
```
1. Do a workspace symbol search
**Expected behavior:**
`Duck` is returned in `navto ` response
**Actual behavior:**
No result for `Duck` included
```
[Trace - 5:52:43 PM] <semantic> Sending request: navto (103). Response expected: yes. Current queue length: 0
Arguments: {
"file": "/Users/matb/projects/san/main.js",
"searchValue": "Duck"
}
[Trace - 5:52:43 PM] <semantic> Response received: navto (103). Request took 49 ms. Success: true
Result: []
```
| Suggestion,Experience Enhancement | low | Critical |
493,096,317 | terminal | Audit uses of the screen buffer/alternate buffer, or come up with a design that makes this not an issue | Is this indicative of a larger problem? Should we file an outstanding work item to audit all sites of manipulating the active buffer and/or somehow change the design such that the methods cannot accidentally be called without resolving which one is the active one first?
For the latter, my proposal would be to hide all the verbage methods off the base class and implement some sort of interface that holds them that is only returned when/if someone has correctly queried for the active one first. I'd accept competing proposals, though.
_Originally posted by @miniksa in https://github.com/microsoft/terminal/pull/2729#issuecomment-530923095_ | Product-Conhost,Area-VT,Issue-Task | low | Major |
493,100,853 | flutter | Benchmark complex_layout_scroll_perf__memory needs better metrics to avoid reporting phantom regressions | The main issue (after some discussion) is that this benchmark is not reporting metrics that exactly match its concern and so situations such as those described below can cause it to indicate a regression when a change simply affects its underlying assumptions instead. We need to develop better metrics for it to measure so that it only flags when we have an actual memory regression.
Until then the benchmark acts mostly as a "canary" that will often see a change that might negatively affect app performance, but sometimes in cases that don't really impact real-world performance. Changes in the benchmark should be noted, but more investigation will determine if the change in reported metrics really represents a regression.
**Originally reported issue:**
I'm investigating the historic slowdowns in the complex_layout_scroll_perf__memory benchmark. It looks like we were at around 5000 in late July and since then we've been stuck around 9000+.
It appears that commit 9b150f134b9b6ba103ec305f489645b5fee905d2 is responsible for that change. Here are 3 runs of the benchmark on a Moto G4 first running on commit 0d0af31598f5ee0552529bfd3fdc35be1d8fe176 (which is just before the indicated commit) and then on commit 9b150f134b9b6ba103ec305f489645b5fee905d2
```
Stats for hash 0d0af3
On G4:
"data": {
"start-min": 34210,
"start-max": 40838,
"start-median": 39220,
"end-min": 44052,
"end-max": 46190,
"end-median": 45034,
"diff-min": 4464,
"diff-max": 11980,
"diff-median": 5033
},
"data": {
"start-min": 35164,
"start-max": 41233,
"start-median": 40473,
"end-min": 44973,
"end-max": 45653,
"end-median": 45321,
"diff-min": 4125,
"diff-max": 10322,
"diff-median": 4487
},
"data": {
"start-min": 32907,
"start-max": 40356,
"start-median": 39693,
"end-min": 44393,
"end-max": 48066,
"end-median": 44879,
"diff-min": 4099,
"diff-max": 15159,
"diff-median": 4907
},
Stats for hash 9b150f:
On G4:
"data": {
"start-min": 31106,
"start-max": 33471,
"start-median": 31184,
"end-min": 40494,
"end-max": 45019,
"end-median": 40980,
"diff-min": 7793,
"diff-max": 11548,
"diff-median": 9319
},
"data": {
"start-min": 31144,
"start-max": 32886,
"start-median": 31240,
"end-min": 40506,
"end-max": 41104,
"end-median": 40906,
"diff-min": 7932,
"diff-max": 9798,
"diff-median": 9690
},
"data": {
"start-min": 30958,
"start-max": 32838,
"start-median": 31365,
"end-min": 40332,
"end-max": 41224,
"end-median": 40892,
"diff-min": 8054,
"diff-max": 9874,
"diff-median": 9501
},
``` | team,engine,c: performance,perf: memory,P3,team-engine,triaged-engine | low | Major |
493,117,649 | TypeScript | Type variables in TypedPropertyDescriptor cause compilation errors when using decorator | **TypeScript Version:** 3.6.3
<!-- Search terms you tried before logging this (so others can find this issue more easily) -->
**Search Terms:** TypedPropertyDescriptor, MethodDecorator, TS2345, Types of property 'value' are incompatible.
**Code**
```ts
type MyDynamicDescriptor = <T>(input: T) => T
type MyStaticDescriptor = (input: number) => number
function dynamicDecorator(foo: string) {
return (_: any, __: any, descriptor: TypedPropertyDescriptor<MyDynamicDescriptor>) => {
console.log('foo:', foo)
console.log('descriptor.value:', descriptor.value)
}
}
function staticDecorator(foo: string) {
return (_: any, __: any, descriptor: TypedPropertyDescriptor<MyStaticDescriptor>) => {
console.log('foo:', foo)
console.log('descriptor.value:', descriptor.value)
}
}
class myClass {
@dynamicDecorator('bar')
myMethod(input: number): number {
return input + 42
}
@staticDecorator('bar')
myOtherMethod(input: number): number {
return input + 42
}
}
```
**Expected behavior:**
No compilation errors for both `dynamicDecorator` and `staticDecorator`.
**Actual behavior:**
Compilation errors for `dynamicDecorator`:
```
error TS2345: Argument of type 'TypedPropertyDescriptor<(input: number) => Promise<number>>' is not assignable to parameter of type 'TypedPropertyDescriptor<MyDynamicDescriptor>'.
Types of property 'value' are incompatible.
Type '((input: number) => Promise<number>) | undefined' is not assignable to type 'MyDynamicDescriptor | undefined'.
Type '(input: number) => Promise<number>' is not assignable to type 'MyDynamicDescriptor'.
```
[**Playground Link**](http://www.typescriptlang.org/play/?experimentalDecorators=true#code/C4TwDgpgBAsiAiIB2BDAtgSwMbwgZywCcMxgB7QqAXigB4AVAPgAoMkwBXYALinoEpqjPgChQkWCADKwFMGy4CxUhWpRW7LryQc0AIwiFBVYTv2GRIgGYckWeWSRQAJsnQKIWCnIrMrZMl48YGIkAHNBAG8RKChCCGAOQidmAH1eFCQQABooVPSoTJyXfCISckJeenAIZwAFQjJIQlBFMpVCWjhEVEwcUuUKxmNhaNjYryQ8MgAbCAA6GbIw5gByf0DV3I3+GPHJ6bnF5bXnAfKKeYA3FBmOCG4tkqULwmvb+93YgF8RX+tbPYMI4oME5B4vIQfIQ-AEgiE2BEoGM4gkkikCkVcvkMllcmcXh0qjV6o1mq1zh0utJZPJ+oShiNkXsJo5Dgslit1nCnjsWVADrMOSdVgT2hV3ncHk8xYNLjcpV8oL9-lgZig8HgoGgQABhdWa5mxAACrl6EO8FTWehQhFWSp1MASAAsyM4NJweFAzAYjNpdL6jeN4olklA2J6oABqKAAFgATHt-iawXTcJDodbbfa9jqAPLAZ2GJ2Ft0erTegOGfj+8yUFGxEPo8OaYDRuOJn5-ERAA) | Needs Investigation | low | Critical |
493,142,300 | flutter | Image loading can make widget tests flakey | We had a widget test that succeeded when run in isolation, but failed when run with others in a group.
This was because we had an image without an explicit size. If the image had been previously loaded in another test and so was cached in memory, the image widget received the correct sizing.
However if the test was run in isolation, the image wasn’t loaded in time, making the layout different.
This was very difficult to debug.
Possible fixes:
1. I would expect that `pumpAndSettle` would wait for images to load and the layout to be complete, this doesn’t seem to be the case.
2. Flutter could clear the in-memory cache at the start of every widget test: `imageCache.clear()`
<!-- Finally, paste the output of running `flutter doctor -v` here. -->
```
[✓] Flutter (Channel stable, v1.9.1+hotfix.2, on Mac OS X 10.14.6 18G95, locale en-GB)
• Flutter version 1.9.1+hotfix.2 at /Users/tom/dev/flutter
• Framework revision 2d2a1ffec9 (6 days ago), 2019-09-06 18:39:49 -0700
• Engine revision b863200c37
• Dart version 2.5.0
[✓] Android toolchain - develop for Android devices (Android SDK version 28.0.3)
• Android SDK at /Users/tom/Library/Android/sdk
• Android NDK location not configured (optional; useful for native profiling support)
• Platform android-28, build-tools 28.0.3
• ANDROID_HOME = /Users/tom/Library/Android/sdk
• Java binary at: /Applications/Android Studio.app/Contents/jre/jdk/Contents/Home/bin/java
• Java version OpenJDK Runtime Environment (build 1.8.0_202-release-1483-b49-5587405)
• All Android licenses accepted.
[✓] Xcode - develop for iOS and macOS (Xcode 10.3)
• Xcode at /Applications/Xcode.app/Contents/Developer
• Xcode 10.3, Build version 10G8
• CocoaPods version 1.7.1
[✓] Android Studio (version 3.5)
• Android Studio at /Applications/Android Studio.app/Contents
• Flutter plugin version 39.0.3
• Dart plugin version 191.8423
• Java version OpenJDK Runtime Environment (build 1.8.0_202-release-1483-b49-5587405)
[✓] VS Code (version 1.38.1)
• VS Code at /Applications/Visual Studio Code.app/Contents
• Flutter extension version 3.4.1
``` | a: tests,framework,a: images,c: proposal,P3,team-framework,triaged-framework | low | Critical |
493,159,553 | TypeScript | Allow overrides in tsconfig | <!-- 🚨 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 -->
- tsconfig
- overrides
- eslint
- glob
## Suggestion
<!-- A summary of what you'd like to see added or changed -->
It would be nice if `tsconfig.json` files would allow an `overrides` section which works in the same way [as ESLint](https://eslint.org/docs/user-guide/configuring#configuration-based-on-glob-patterns). Here I have just one config file in my root for all kind of files and use cases. TypeScript currently needs multiple `tsconfig.json` files.
## Use Cases
<!--
What do you want to use this for?
What shortcomings exist with current approaches?
-->
As far as I know editors like VS Code look for the nearest `tsconfig.json` for proper TypeScript support. This make it hard to properly type files which are used for different environments (e.g. source code vs tests vs storybook examples and so on) in the same directory, which seems to be a common convention nowadays (e.g. `src/file.ts`, `src/file.test.ts`, `src/file.stories.ts`).
Most people seem to ignore this fact which makes test globals like `describe` available in source code files.
I once [tweeted](https://twitter.com/PipoPeperoni/status/1166228487268683777
) about this with a small example:
## Examples
<!-- Show how this would be used and what the behavior would be -->
```json
{
"compilerOptions": {
"allowSyntheticDefaultImports": true,
"module": "esnext",
"moduleResolution": "node",
"noEmit": true,
"strict": true,
"skipLibCheck": true,
"target": "es2017",
"types": []
},
"overrides": [
{
"include": ["src/**/*"],
"exclude": ["src/**/*.test.ts"],
"compilerOptions": {
"jsx": "react",
"types": ["webpack-env"],
"plugins": [
{
"name": "typescript-styled-plugin"
}
]
}
},
{
"include": ["src/**/*.test.ts"],
"compilerOptions": {
"types": ["jest", "node"]
}
}
]
}
```
## 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 | high | Critical |
493,244,750 | three.js | Editor : Enhancments and Features addition | ##### Editor Enhancments
Creating this thread to track implementations for the editor enhancments and features like Material Browser.
- [x] TabbedPanel UI Element #17465.
- [x] Listbox UI Element. #17490
- [x] MaterialBrowser Panel in project panel with menu to Create, And Apply Material and store it in the editor. #17556
- [ ] Open the Selected Material ( in MaterialBrowser ) in Properties editor.
- [ ] Replace the current History panel with the Listbox Element.
- [ ] Refactor Outliner to create a generic simple treeview.
- [ ] Replace the current Outliner for Scene Panel with the new one.
- [ ] Cleanup.
Now that we are implementing this i propose to refactor the Add* commands to not apply a default material while creating objects and adding to the scene. | Editor | low | Minor |
493,270,095 | flutter | Enable touch filtering and tapjacking protection |
## Use case
Currently, there is no way to enable touch filtering in buttons like the android feature **setFilterTouchesWhenObscured(true).** This is crucial to protect apps from tapjacking.
## Proposal
A property on ButtonWidgets that will enable touch filtering :
`
FlatButton(
child: Text(Strings.ok),
enableFilterTouchesWhenObscured(true)
onPressed: () {
Navigator.of(context).pop();
},
)
` | c: new feature,framework,P3,team-framework,triaged-framework | medium | Critical |
493,308,770 | pytorch | Memory leak in multithreading environment when loading checkpoint | ## 🐛 Bug
I have a problem when loading saved checkpoint for pytorch model in seperate thread. CPU memory keeps increasing and is never released.
Multi-threading version of script increases RAM usage with each iteration and ends with 978 RESV memory (htop output).
Single-threading version holds with 374 RESV memory (htop output).
When I randomly initialize model (without state_dict loading) both versions use the same (smaller) amount of memory.
## To Reproduce
```
import torch
import time
import threading
import torchvision
def create_tensor(index):
print("Create model {}".format(index))
resnet50 = torchvision.models.resnet50(pretrained=True)
del resnet50
def create_tensor_thread(index):
x = threading.Thread(target=create_tensor, args=(index,))
x.start()
x.join()
def run(use_threading=False):
for index in range(5):
if use_threading:
#memory leak observed
create_tensor_thread(index)
else:
#no memory leak observed
create_tensor(index)
time.sleep(5)
run(True)
```
## Expected behavior
I expect the RAM to be cleared as no reference is being held to created models.
## Additional info
I have traced the issue to https://github.com/pytorch/pytorch/blob/33221b19acc3dcacb11c38fdbff65d9a6ce90866/torch/nn/modules/module.py#L775
Commenting out this line makes the multi-threading version behave as single-threading one.
## Environment
- PyTorch Version: 1.2.0
- OS: Ubuntu 18.04
- How you installed PyTorch: pip
- Python version: 3.5.7
cc @ezyang @gchanan @zou3519 @jerryzh168 | high priority,needs reproduction,module: multiprocessing,module: memory usage,triaged,module: multithreading | low | Critical |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.