Implementation commitments on RTCEncodingOptions.keyFrame?

[dontcallmedom]

RTCEncodingOptions.keyFrame (for keyframe generation request) has reasonable test coverage (although in the wrong directory at the moment), one passing implementation.

@jan-ivar @youennf I could not find any signal of implementation commitment from Firefox or webkit on this; is this something on which you could shed some light?

[fippo]

Firefox position and Webkit position, no response on either.

I count two passing implementations?

[dontcallmedom]

I count two passing implementations?

if you're thinking Chrome and Edge, they're not considered independent implementation

[fippo]

How are "independent implementations" defined then? In this case all four browsers listed use the same underlying implementation (libWebRTCs video engine)

[dontcallmedom]

right, and the WG has long acknowledged that this wasn't ideal in terms of independence; but the next best thing is implementations that don't share the same integration layer between the JS API and libwebrtc (which historically has proved productive in identifying bugs / disagreements in the spec)

[fippo]

from libWebRTCs git log --format="%ae" --since=2020-01-01:
Google: 10385 (webrtc.org, chromium.org, google.com)
Microsoft: 294
Mozilla: 46
Apple: 0