- From: Jan-Ivar Bruaroey via GitHub <sysbot+gh@w3.org>
- Date: Thu, 09 Mar 2023 13:27:07 +0000
- To: public-webrtc-logs@w3.org
I see my mistake now. I misread what the [original text](https://w3c.github.io/mediacapture-main/getusermedia.html#dfn-stop-all-sources) was trying to do (I read it as inadvertently stopping all tracks of a physical source in the user agent). But in fixing it, it fails to preserve the original intent, which I failed to interpret and we wish to to preserve, which is effectively to end all tracks "tied" to the `mediaDevices` object they came from (through `mediaDevices.getUserMedia()`), so that even tracks transferred elsewhere end when this document unloads. Or as mediacapture-extensions [says](https://w3c.github.io/mediacapture-extensions/#transferable-mediastreamtrack): _"When originalContext goes away, trackSource gets ended, thus transferredTrack gets ended."_ I'll try to update my PR to not break that. To be fair, I believe the existing language only accomplishes that under ONE interpretation of the scope of source (one per global object), NOT the other (source is a singleton of the user agent), but I believe we can express while being agnostic of source scope. > I think the intent of the spec is clear, as we discussed this when designing transferred streams. We can always clarify this in the spec. This is all I'm trying to do. :) -- GitHub Notification of comment by jan-ivar Please view or discuss this issue at https://github.com/w3c/mediacapture-main/pull/939#issuecomment-1462070034 using your GitHub account -- Sent via github-notify-ml as configured in https://github.com/w3c/github-notify-ml-config
Received on Thursday, 9 March 2023 13:27:09 UTC