Re: WebRTC 1.0 CR exit criteria

Le 05/04/2017 à 19:01, Bernard Aboba a écrit :
> [BA] To me, it does not make sense to retain optional features with
> only one implementation while removing mandatory features with only a
> single implementation.
> 
> Aside from Section 8.6 (where mandatory to implement stats are
> identified), the specification rarely
> calls out optional functionality.
> 
> Can we perhaps start from the assumption that unless called out as
> optional, features are mandatory to implement?

+1

> If a feature only turns out to have one implementation, can we
> reclassify it as optional at PR?

If it was marked at risk, probably; but if not, it "just" means another
CR cycle which can stay pretty short.

Dom

Received on Thursday, 6 April 2017 09:12:06 UTC