Re: [webrtc-pc] Why is RTCRtpSynchronizationSource.voiceActivityFlag required-but-nullable?

So in Nightly we've implemented `getSynchronizationSources()` without implementing the `voiceActivityFlag`, and users can tell because the member will be absent. Had we returned `null` people might mistake it for a working implementation.

Now, you could probably argue we shouldn't have implemented anything until we'd implemented all the fields - you'd have a point, we've been less than great in WebRTC at holding back APIs until they're complete - in practice, I think it's also true that there's always another partial dictionary around the corner, so what you're saying doesn't seem to work in general. E.g. if there were a new similar member spec'ed tomorrow, shipping updated webidl that returned `null` for this member wouldn't seem to make much sense, introducing a js observable change at a point in time of no significance.

What's the rationale for avoiding required-but-null on outputs? 

-- 
GitHub Notification of comment by jan-ivar
Please view or discuss this issue at https://github.com/w3c/webrtc-pc/issues/1689#issuecomment-350865595 using your GitHub account

Received on Monday, 11 December 2017 21:34:33 UTC