1301675 - Allow stopping all MediaStreamTracks

Status: RESOLVED FIXED

(Core :: Audio/Video: MediaStreamGraph, defect, P2)

Description

[Andreas Pehrson [:pehrsons]]

Since implementing MediaStreamTrack.stop, the spec has changed from allowing only stopping of local tracks to all kinds of tracks.

We need to remove MediaStreamTrackSource::mRemote and add support for stopping all different kinds of MediaStreamTrackSources per the following


Canvas captureStream: can stop the capturing completely, also disconnect the track and stream so the canvas can be GCed even with the track or stream alive

Media captureStream: will keep producing new tracks so we'll probably do nothing

WebAudio MediaStreamDestination: can stop the capturing completely, also disconnect the track and stream so the node can be GCed even with the track or stream alive

RTCPeerConnection: A stopped track shouldn't affect the peer connection

Comment 1

[Andreas Pehrson [:pehrsons]]

Attachment: Bug 1301675 - Remove MediaStreamTrackSource::mIsRemote.

MediaStreamTrack.remote is no longer part of the spec.

Review commit: https://reviewboard.mozilla.org/r/80768/diff/#index_header
See other reviews: https://reviewboard.mozilla.org/r/80768/

Comment 2

[Andreas Pehrson [:pehrsons]]

Attachment: Bug 1301675 - Test stopping tracks from canvas captureStream.

Review commit: https://reviewboard.mozilla.org/r/80770/diff/#index_header
See other reviews: https://reviewboard.mozilla.org/r/80770/

Comment 3

[Andreas Pehrson [:pehrsons]]

Attachment: Bug 1301675 - Test stopping tracks from canvas (webgl) captureStream.

Review commit: https://reviewboard.mozilla.org/r/80772/diff/#index_header
See other reviews: https://reviewboard.mozilla.org/r/80772/

Comment 4

[Andreas Pehrson [:pehrsons]]

Attachment: Bug 1301675 - Refactor canvas captureStream to be more clear on removing frame listeners.

Review commit: https://reviewboard.mozilla.org/r/80774/diff/#index_header
See other reviews: https://reviewboard.mozilla.org/r/80774/

Comment 5

[Andreas Pehrson [:pehrsons]]

Attachment: Bug 1301675 - Implement CanvasCaptureTrackSource that allows stopping canvas capture.

Review commit: https://reviewboard.mozilla.org/r/80776/diff/#index_header
See other reviews: https://reviewboard.mozilla.org/r/80776/

Comment 6

[Andreas Pehrson [:pehrsons]]

Attachment: Bug 1301675 - Clean up MediaStream Constructor test.

Review commit: https://reviewboard.mozilla.org/r/80778/diff/#index_header
See other reviews: https://reviewboard.mozilla.org/r/80778/

Comment 7

[Andreas Pehrson [:pehrsons]]

Attachment: Bug 1301675 - Implement StreamCaptureTrackSource::Stop.

Review commit: https://reviewboard.mozilla.org/r/80780/diff/#index_header
See other reviews: https://reviewboard.mozilla.org/r/80780/

Comment 8

[Andreas Pehrson [:pehrsons]]

Attachment: Bug 1301675 - Clarify why we don't need to do anything on DecoderCaptureTrackSource::Stop().

Review commit: https://reviewboard.mozilla.org/r/80782/diff/#index_header
See other reviews: https://reviewboard.mozilla.org/r/80782/

Comment 9

[Andreas Pehrson [:pehrsons]]

Attachment: Bug 1301675 - Implement AudioDestinationTrackSource.

So stop()ing the output track can clean up the destination node.

Review commit: https://reviewboard.mozilla.org/r/80784/diff/#index_header
See other reviews: https://reviewboard.mozilla.org/r/80784/

Comment 10

[Andreas Pehrson [:pehrsons]]

Attachment: Bug 1301675 - Test that a track from MediaStreamAudioDestinationNode can be stopped.

Review commit: https://reviewboard.mozilla.org/r/80786/diff/#index_header
See other reviews: https://reviewboard.mozilla.org/r/80786/

Comment 11

[Andreas Pehrson [:pehrsons]]

Attachment: Bug 1301675 - Received tracks from RTCPeerConnection are stoppable.

Review commit: https://reviewboard.mozilla.org/r/80788/diff/#index_header
See other reviews: https://reviewboard.mozilla.org/r/80788/

Comment 12

[Andreas Pehrson [:pehrsons]]

Attachment: Bug 1301675 - Rename BasicUnstoppableTrackSource to BasicTrackSource.

They are not unstoppable any longer. We just don't forward Stop() to the real source.

Review commit: https://reviewboard.mozilla.org/r/80790/diff/#index_header
See other reviews: https://reviewboard.mozilla.org/r/80790/

Comment 13

[Andreas Pehrson [:pehrsons]]

Attachment: Bug 1301675 - Fix track cloning test.

It relied on track.stop() not working for MediaStreamAudioDestinationNode tracks.

Review commit: https://reviewboard.mozilla.org/r/80792/diff/#index_header
See other reviews: https://reviewboard.mozilla.org/r/80792/

Comment 14

[Andreas Pehrson [:pehrsons]]

Attachment: Bug 1301675 - Uncomment assertion for when removed live track was not found.

It needs an extra guard for when we're an audio element and a live video track
was removed. That's ok and we should ignore it.

Review commit: https://reviewboard.mozilla.org/r/80794/diff/#index_header
See other reviews: https://reviewboard.mozilla.org/r/80794/

Comment 15

[Andreas Pehrson [:pehrsons]]

Attachment: Bug 1301675 - Assert that a MediaStreamTrackSource is not stopped more than once.

Review commit: https://reviewboard.mozilla.org/r/80796/diff/#index_header
See other reviews: https://reviewboard.mozilla.org/r/80796/

Comment 16

[Paul Adenot (:padenot)]

Comment on attachment 8794267 [details]
Bug 1301675 - Implement AudioDestinationTrackSource.

https://reviewboard.mozilla.org/r/80784/#review79656

Comment 17

[Paul Adenot (:padenot)]

Comment on attachment 8794268 [details]
Bug 1301675 - Test that a track from MediaStreamAudioDestinationNode can be stopped.

https://reviewboard.mozilla.org/r/80786/#review79658

Comment 18

[Jan-Ivar Bruaroey [:jib] (needinfo? me)]

(In reply to Andreas Pehrson [:pehrsons] (Telenor) from comment #0)
> RTCPeerConnection: A stopped track shouldn't affect the peer connection

Actually, track.stop() has been overloaded to be how apps reject an offered track, both during offer-answer negotiation [1], and after [2].

[1] http://w3c.github.io/webrtc-pc/#processing-remote-mediastreamtracks
[2] http://w3c.github.io/webrtc-pc/#setting-negotiation-needed

Comment 19

[Jan-Ivar Bruaroey [:jib] (needinfo? me)]

See also http://w3c.github.io/webrtc-pc/#dom-rtcrtpreceiver-track

Comment 20

[Jan-Ivar Bruaroey [:jib] (needinfo? me)]

Comment on attachment 8794259 [details]
Bug 1301675 - Remove MediaStreamTrackSource::mIsRemote.

https://reviewboard.mozilla.org/r/80768/#review81248

Comment 21

[Jan-Ivar Bruaroey [:jib] (needinfo? me)]

Comment on attachment 8794260 [details]
Bug 1301675 - Test stopping tracks from canvas captureStream.

https://reviewboard.mozilla.org/r/80770/#review81252

::: dom/canvas/test/test_capture.html:111
(Diff revision 1)
> +        ok(true, "Element " + elem.id + " ended.");
> +        resolve();
> +        elem.removeEventListener("ended", endedListener);
> +      }));
> +  });
> +  return Promise.all(promises);

promises could be inlined...

Comment 22

[Jan-Ivar Bruaroey [:jib] (needinfo? me)]

Comment on attachment 8794261 [details]
Bug 1301675 - Test stopping tracks from canvas (webgl) captureStream.

https://reviewboard.mozilla.org/r/80772/#review81254

::: dom/canvas/test/webgl-mochitest/test_capture.html:121
(Diff revision 1)
> +    return new Promise(resolve =>
> +      elem.addEventListener("ended", function endedListener(event) {
> +        ok(true, "Element " + elem.id + " ended.");
> +        resolve();
> +        elem.removeEventListener("ended", endedListener);
> +      }));

It'd be nice to move haveEvent() someplace more central, but where?

Comment 23

[Jan-Ivar Bruaroey [:jib] (needinfo? me)]

Comment on attachment 8794262 [details]
Bug 1301675 - Refactor canvas captureStream to be more clear on removing frame listeners.

https://reviewboard.mozilla.org/r/80774/#review81256

Comment 24

[Jan-Ivar Bruaroey [:jib] (needinfo? me)]

Comment on attachment 8794263 [details]
Bug 1301675 - Implement CanvasCaptureTrackSource that allows stopping canvas capture.

https://reviewboard.mozilla.org/r/80776/#review81262

::: dom/html/HTMLCanvasElement.cpp:688
(Diff revision 1)
> +NS_IMPL_ADDREF_INHERITED(CanvasCaptureTrackSource,
> +                         MediaStreamTrackSource)
> +NS_IMPL_RELEASE_INHERITED(CanvasCaptureTrackSource,
> +                          MediaStreamTrackSource)

Any reason not to use NS_IMPL_ISUPPORTS_INHERITED ?

Comment 25

[Jan-Ivar Bruaroey [:jib] (needinfo? me)]

Comment on attachment 8794264 [details]
Bug 1301675 - Clean up MediaStream Constructor test.

https://reviewboard.mozilla.org/r/80778/#review81264

Comment 26

[Jan-Ivar Bruaroey [:jib] (needinfo? me)]

Comment on attachment 8794265 [details]
Bug 1301675 - Implement StreamCaptureTrackSource::Stop.

https://reviewboard.mozilla.org/r/80780/#review81266

::: dom/html/HTMLMediaElement.cpp:1422
(Diff revision 1)
> +    if (!ms.mCapturingMediaStream) {
> +      continue;
> +    }
> +
> +    if (ms.mStream != aOwningStream) {
> +      continue;
> +    }

combine?

::: dom/html/HTMLMediaElement.cpp:1430
(Diff revision 1)
> +    for (int32_t i = ms.mTrackPorts.Length() - 1; i >= 0; --i) {
> +      MediaInputPort* port = ms.mTrackPorts[i].second();
> +      if (port->GetDestinationTrackId() != aDestinationTrackID) {
> +        continue;
> +      }
> +
> +      port->Destroy();
> +      ms.mTrackPorts.RemoveElementAt(i);
> +      return;

Any reason this needs to be iterated in reverse?

::: dom/html/HTMLMediaElement.cpp:2266
(Diff revision 1)
> -  explicit StreamCaptureTrackSource(MediaStreamTrackSource* aCapturedTrackSource)
> +  StreamCaptureTrackSource(HTMLMediaElement* aElement,
> +                           MediaStreamTrackSource* aCapturedTrackSource,
> +                           DOMMediaStream* aOwningStream,
> +                           TrackID aDestinationTrackID)

Why implicit?

Comment 27

[Jan-Ivar Bruaroey [:jib] (needinfo? me)]

Comment on attachment 8794266 [details]
Bug 1301675 - Clarify why we don't need to do anything on DecoderCaptureTrackSource::Stop().

https://reviewboard.mozilla.org/r/80782/#review81270

Comment 28

[Jan-Ivar Bruaroey [:jib] (needinfo? me)]

Comment on attachment 8794269 [details]
Bug 1301675 - Received tracks from RTCPeerConnection are stoppable.

https://reviewboard.mozilla.org/r/80788/#review81272

::: media/webrtc/signaling/src/peerconnection/PeerConnectionMedia.h:177
(Diff revision 1)
>    ApplyConstraints(nsPIDOMWindowInner* aWindow,
>                     const dom::MediaTrackConstraints& aConstraints) override;
>  
>    void Stop() override
>    {
> -    // XXX Fix in later patch.
> +    // No need to do anything when stopping a remote track.

See comment 18. Update code comment if nothing else.

Comment 29

[Jan-Ivar Bruaroey [:jib] (needinfo? me)]

Comment on attachment 8794270 [details]
Bug 1301675 - Rename BasicUnstoppableTrackSource to BasicTrackSource.

https://reviewboard.mozilla.org/r/80790/#review81276

Comment 30

[Jan-Ivar Bruaroey [:jib] (needinfo? me)]

Comment on attachment 8794271 [details]
Bug 1301675 - Fix track cloning test.

https://reviewboard.mozilla.org/r/80792/#review81278

Comment 31

[Jan-Ivar Bruaroey [:jib] (needinfo? me)]

Comment on attachment 8794272 [details]
Bug 1301675 - Uncomment assertion for when removed live track was not found.

https://reviewboard.mozilla.org/r/80794/#review81280

::: dom/html/HTMLMediaElement.cpp:4307
(Diff revision 1)
> -    // XXX (bug 1208328) Uncomment this when DOMMediaStream doesn't call
> -    // NotifyTrackRemoved multiple times for the same track, i.e., when it
> +    NS_ASSERTION(aTrack->AsVideoStreamTrack() && !IsVideo(),
> +                 "MediaStreamTrack ended but did not exist in track lists");

Please explain this assertion.

Comment 32

[Jan-Ivar Bruaroey [:jib] (needinfo? me)]

Comment on attachment 8794273 [details]
Bug 1301675 - Assert that a MediaStreamTrackSource is not stopped more than once.

https://reviewboard.mozilla.org/r/80796/#review81284

Comment 33

[Andreas Pehrson [:pehrsons]]

Comment on attachment 8794263 [details]
Bug 1301675 - Implement CanvasCaptureTrackSource that allows stopping canvas capture.

https://reviewboard.mozilla.org/r/80776/#review81262

> Any reason not to use NS_IMPL_ISUPPORTS_INHERITED ?

I don't know these macros enough. But it seems to me like NS_IMPL_ISUPPORTS_INHERITED and friends don't give us the NS_INTERFACE macro that we need. We need NS_INTERFACE_MAP_BEGIN_CYCLE_COLLECTION_INHERITED. And they can't be combined.

Comment 34

[Andreas Pehrson [:pehrsons]]

Comment on attachment 8794265 [details]
Bug 1301675 - Implement StreamCaptureTrackSource::Stop.

https://reviewboard.mozilla.org/r/80780/#review81266

> combine?

I find it cleaner to separate the guards. Especially when you want to add log statements to see which of them are guarding the call off - even if it's just a temporary local debug statement.

> Any reason this needs to be iterated in reverse?

`ms.mTrackPorts.RemoveElementAt(i)` wouldn't work if done in a regular loop.

> Why implicit?

Explicit only comes into effect when there's exactly one argument to the constructor, no?

Comment 35

[Jan-Ivar Bruaroey [:jib] (needinfo? me)]

Comment on attachment 8794265 [details]
Bug 1301675 - Implement StreamCaptureTrackSource::Stop.

https://reviewboard.mozilla.org/r/80780/#review81266

> Explicit only comes into effect when there's exactly one argument to the constructor, no?

Yeah, D'oh!

Comment 36

[Andreas Pehrson [:pehrsons]]

(In reply to Jan-Ivar Bruaroey [:jib] from comment #28)
> Comment on attachment 8794269 [details]
> Bug 1301675 - Received tracks from RTCPeerConnection are stoppable.
> 
> https://reviewboard.mozilla.org/r/80788/#review81272
> 
> ::: media/webrtc/signaling/src/peerconnection/PeerConnectionMedia.h:177
> (Diff revision 1)
> >    ApplyConstraints(nsPIDOMWindowInner* aWindow,
> >                     const dom::MediaTrackConstraints& aConstraints) override;
> >  
> >    void Stop() override
> >    {
> > -    // XXX Fix in later patch.
> > +    // No need to do anything when stopping a remote track.
> 
> See comment 18. Update code comment if nothing else.

I can't tell from the links in comment 18 and 19 what is supposed to happen when a received track is stopped. Can you point this out in the spec, or should we wait until it becomes more clear? It's a recent decision I take it.

Comment 37

[Jan-Ivar Bruaroey [:jib] (needinfo? me)]

(In reply to Andreas Pehrson [:pehrsons] (Telenor) from comment #36)
> I can't tell from the links in comment 18 and 19 what is supposed to happen
> when a received track is stopped. Can you point this out in the spec, or
> should we wait until it becomes more clear? It's a recent decision I take it.

It's not that recent, but not particularly clear either. From running this fiddle [1] in Firefox you can see that ontrack events fire during setRemoteDescription. From reading earlier versions of the spec [2] it seems the intent is for JS to call track.stop() from within an ontrack callback to "reject" that track, which I think has an effect on the SDP negotiation at that point, or at least allocated bandwidth.

Comment 19 suggests more generally that track.stop() on a remote track even after negotiation causes it at least to stop playback. It says "Since receiver.track.stop() does not implicitly stop receiver, Receiver Reports continue to be sent.", so that may be it. I see no mention of this setting the negotiationneeded flag...

Except the second link in comment 18 suggests that stopping a track sets negotiationneeded, but it doesn't say whether this is sender side, receiver side or both.

CCing Byron to see if he knows.

Looking at Chrome for clues: Chrome doesn't have ontrack yet, so when I run the fiddle [1] in Chrome with adapter.js, you'll note they fire after setRemoteDescription (this is possibly caused by adapter's ontrack polyfill), so nothing happens to the SDP AFAICT.

Note that the fiddle calls track.stop() on the video, and in Chrome, this causes the remote video to not appear. Stopping audio instead didn't work though, so we may have some time here.

[1] https://jsfiddle.net/c2vwpuqg/
[1] https://github.com/w3c/webrtc-pc/commit/092504a8276dee977c4710666b3203f0d853736b

Comment 38

[Andreas Pehrson [:pehrsons]]

(In reply to Jan-Ivar Bruaroey [:jib] from comment #37)
> Comment 19 suggests more generally that track.stop() on a remote track even
> after negotiation causes it at least to stop playback. It says "Since
> receiver.track.stop() does not implicitly stop receiver, Receiver Reports
> continue to be sent.", so that may be it. I see no mention of this setting
> the negotiationneeded flag...

This is what I thought would happen, and was agreed upon at some point. That the tracks stops but we don't have to tell anyone.

Most of the discussion seems to be here: https://github.com/w3c/webrtc-pc/issues/597

I don't see anyone mentioning track.stop() having a different effect if called during negotiation.

Comment 39

[Byron Campen [:bwc]]

I can't resolve the ambiguity here, I'm afraid.

Comment 40

[Andreas Pehrson [:pehrsons]]

I raised an issue with the spec at https://github.com/w3c/webrtc-pc/issues/898.

Comment 41

[Andreas Pehrson [:pehrsons]]

Comment on attachment 8794259 [details]
Bug 1301675 - Remove MediaStreamTrackSource::mIsRemote.

Review request updated; see interdiff: https://reviewboard.mozilla.org/r/80768/diff/1-2/

Comment 42

[Andreas Pehrson [:pehrsons]]

Comment on attachment 8794260 [details]
Bug 1301675 - Test stopping tracks from canvas captureStream.

Review request updated; see interdiff: https://reviewboard.mozilla.org/r/80770/diff/1-2/

Comment 43

[Andreas Pehrson [:pehrsons]]

Comment on attachment 8794261 [details]
Bug 1301675 - Test stopping tracks from canvas (webgl) captureStream.

Review request updated; see interdiff: https://reviewboard.mozilla.org/r/80772/diff/1-2/

Comment 44

[Andreas Pehrson [:pehrsons]]

Comment on attachment 8794262 [details]
Bug 1301675 - Refactor canvas captureStream to be more clear on removing frame listeners.

Review request updated; see interdiff: https://reviewboard.mozilla.org/r/80774/diff/1-2/

Comment 45

[Andreas Pehrson [:pehrsons]]

Comment on attachment 8794263 [details]
Bug 1301675 - Implement CanvasCaptureTrackSource that allows stopping canvas capture.

Review request updated; see interdiff: https://reviewboard.mozilla.org/r/80776/diff/1-2/

Comment 46

[Andreas Pehrson [:pehrsons]]

Comment on attachment 8794264 [details]
Bug 1301675 - Clean up MediaStream Constructor test.

Review request updated; see interdiff: https://reviewboard.mozilla.org/r/80778/diff/1-2/

Comment 47

[Andreas Pehrson [:pehrsons]]

Comment on attachment 8794265 [details]
Bug 1301675 - Implement StreamCaptureTrackSource::Stop.

Review request updated; see interdiff: https://reviewboard.mozilla.org/r/80780/diff/1-2/

Comment 48

[Andreas Pehrson [:pehrsons]]

Comment on attachment 8794266 [details]
Bug 1301675 - Clarify why we don't need to do anything on DecoderCaptureTrackSource::Stop().

Review request updated; see interdiff: https://reviewboard.mozilla.org/r/80782/diff/1-2/

Comment 49

[Andreas Pehrson [:pehrsons]]

Comment on attachment 8794267 [details]
Bug 1301675 - Implement AudioDestinationTrackSource.

Review request updated; see interdiff: https://reviewboard.mozilla.org/r/80784/diff/1-2/

Comment 50

[Andreas Pehrson [:pehrsons]]

Comment on attachment 8794268 [details]
Bug 1301675 - Test that a track from MediaStreamAudioDestinationNode can be stopped.

Review request updated; see interdiff: https://reviewboard.mozilla.org/r/80786/diff/1-2/

Comment 51

[Andreas Pehrson [:pehrsons]]

Comment on attachment 8794269 [details]
Bug 1301675 - Received tracks from RTCPeerConnection are stoppable.

Review request updated; see interdiff: https://reviewboard.mozilla.org/r/80788/diff/1-2/

Comment 52

[Andreas Pehrson [:pehrsons]]

Comment on attachment 8794270 [details]
Bug 1301675 - Rename BasicUnstoppableTrackSource to BasicTrackSource.

Review request updated; see interdiff: https://reviewboard.mozilla.org/r/80790/diff/1-2/

Comment 53

[Andreas Pehrson [:pehrsons]]

Comment on attachment 8794271 [details]
Bug 1301675 - Fix track cloning test.

Review request updated; see interdiff: https://reviewboard.mozilla.org/r/80792/diff/1-2/

Comment 54

[Andreas Pehrson [:pehrsons]]

Comment on attachment 8794272 [details]
Bug 1301675 - Uncomment assertion for when removed live track was not found.

Review request updated; see interdiff: https://reviewboard.mozilla.org/r/80794/diff/1-2/

Comment 55

[Andreas Pehrson [:pehrsons]]

Comment on attachment 8794273 [details]
Bug 1301675 - Assert that a MediaStreamTrackSource is not stopped more than once.

Review request updated; see interdiff: https://reviewboard.mozilla.org/r/80796/diff/1-2/

Comment 56

[Jan-Ivar Bruaroey [:jib] (needinfo? me)]

Comment on attachment 8794269 [details]
Bug 1301675 - Received tracks from RTCPeerConnection are stoppable.

https://reviewboard.mozilla.org/r/80788/#review89272

Comment 57

[Pulsebot]

Pushed by [e-mail]:
https://hg.mozilla.org/integration/mozilla-inbound/rev/7414623c1788
Remove MediaStreamTrackSource::mIsRemote. r=jib
https://hg.mozilla.org/integration/mozilla-inbound/rev/6000b9799462
Test stopping tracks from canvas captureStream. r=jib
https://hg.mozilla.org/integration/mozilla-inbound/rev/d06f0960937f
Test stopping tracks from canvas (webgl) captureStream. r=jib
https://hg.mozilla.org/integration/mozilla-inbound/rev/39cdaa7a776e
Refactor canvas captureStream to be more clear on removing frame listeners. r=jib
https://hg.mozilla.org/integration/mozilla-inbound/rev/e7d21ef28d04
Implement CanvasCaptureTrackSource that allows stopping canvas capture. r=jib
https://hg.mozilla.org/integration/mozilla-inbound/rev/2f28fe5a7449
Clean up MediaStream Constructor test. r=jib
https://hg.mozilla.org/integration/mozilla-inbound/rev/b01b14b70c58
Implement StreamCaptureTrackSource::Stop. r=jib
https://hg.mozilla.org/integration/mozilla-inbound/rev/e31e0a89ecce
Clarify why we don't need to do anything on DecoderCaptureTrackSource::Stop(). r=jib
https://hg.mozilla.org/integration/mozilla-inbound/rev/945dfdabefaa
Implement AudioDestinationTrackSource. r=padenot
https://hg.mozilla.org/integration/mozilla-inbound/rev/f64665a7edc7
Test that a track from MediaStreamAudioDestinationNode can be stopped. r=padenot
https://hg.mozilla.org/integration/mozilla-inbound/rev/b55b83b03e87
Received tracks from RTCPeerConnection are stoppable. r=jib
https://hg.mozilla.org/integration/mozilla-inbound/rev/2e5dc6155ced
Rename BasicUnstoppableTrackSource to BasicTrackSource. r=jib
https://hg.mozilla.org/integration/mozilla-inbound/rev/2a7dfea97b90
Fix track cloning test. r=jib
https://hg.mozilla.org/integration/mozilla-inbound/rev/a9ffc03664bf
Uncomment assertion for when removed live track was not found. r=jib
https://hg.mozilla.org/integration/mozilla-inbound/rev/8e63884e4c09
Assert that a MediaStreamTrackSource is not stopped more than once. r=jib

Comment 58

[Phil Ringnalda (:philor)]

https://hg.mozilla.org/mozilla-central/rev/7414623c1788
https://hg.mozilla.org/mozilla-central/rev/6000b9799462
https://hg.mozilla.org/mozilla-central/rev/d06f0960937f
https://hg.mozilla.org/mozilla-central/rev/39cdaa7a776e
https://hg.mozilla.org/mozilla-central/rev/e7d21ef28d04
https://hg.mozilla.org/mozilla-central/rev/2f28fe5a7449
https://hg.mozilla.org/mozilla-central/rev/b01b14b70c58
https://hg.mozilla.org/mozilla-central/rev/e31e0a89ecce
https://hg.mozilla.org/mozilla-central/rev/945dfdabefaa
https://hg.mozilla.org/mozilla-central/rev/f64665a7edc7
https://hg.mozilla.org/mozilla-central/rev/b55b83b03e87
https://hg.mozilla.org/mozilla-central/rev/2e5dc6155ced
https://hg.mozilla.org/mozilla-central/rev/2a7dfea97b90
https://hg.mozilla.org/mozilla-central/rev/a9ffc03664bf
https://hg.mozilla.org/mozilla-central/rev/8e63884e4c09

Comment 59

[Eric Shepherd [:sheppy]]

I’ve updated the documentation for MediaStreamTrack.stop(), hopefully accurately. I would appreciate a quick look at it to be sure it’s accurate.

The following pages have changed:

https://developer.mozilla.org/en-US/docs/Web/API/MediaStreamTrack/stop

With a mention on Firefox 52 for developers.

Comment 60

[Andreas Pehrson [:pehrsons]]

Thanks! I took a look and made some fixes, though I'm not sure if they affected your changes or older text.