Closed fippo closed 2 years ago
@alvestrand @henbos there is an interesting thing about this sample that I only noticed when writing the test: signaling state and ice connection state get a state change to "new/stable" (from "new/stabe"), connectionState does not. That seems inconsistent...
Yeah I think we should calculate connectionState the same way we calculate iceConnectionState, in blink, so that it matches the JS exposed transport states.
But incidentally connectionState still exposes the ”failed” heuristic that was removed from iceConnectionState.
My kingdom for an end-of-candidates
ah... turns out there was an extra call adding the 'new' / 'stable' state to the div. Not a spec issue, a sample issue. Updated the sample and the test.
and fix misc bugs and terminology