This is an archived snapshot of W3C's public bugzilla bug tracker, decommissioned in April 2019. Please see the home page for more details.
http://dev.w3.org/2011/webrtc/editor/webrtc.html#rtcicecandidate-type It's not clear what the browser's obligations are with respect to sdpMid and sdpMLineIndex. There is an "If present" attached to sdpMid, which might imply something. I think that we should always, always include both in events. And we should require that input include one or other. (And we should work out which takes precedence, or surface an error if both are present.) Did we agree to remove the whole null candidate thing by surfacing an ICE state change?
We did agree to surface an ICE state change. I think we need to keep onicecandidate(null) for backcompat though.
Plan is to make a PR to add onicegatheringstatechange and leave onicecandidate(null)
WebRTC implementations will always generate MID. (if the JSEP docs don't already say so, they should be fixed.) There should be a separate bug for the icegatheringstatechange event.
https://github.com/w3c/webrtc-pc/pull/27 As it turns out onicecandidate(null) wasn't documented, or even permitted.
WebRTC API bugs have been moved to github issues: https://github.com/w3c/webrtc-pc/issues Please subscribe to the issues you want to keep watching.