On 12/22/10 07:18, Bernard Aboba wrote:
It does appear presumptive to suggest that a codec that hasn't completed a standardization process be made "mandatory to implement."
Since there have been some large judgments over use of allegedly "free" codecs, the lesson is that codecs that are claimed to be "free of encumbrance" may in time be discovered not to be. The IETF process can potentially be useful in helping to clarify the IPR status of codecs. However, those wheels grind slowly. I agree that we can't make anything mandatory to implement that we don't have an accepted stable, publicly available reference for. (I'm working on solving that for the case of VP8).
However, I don't agree that we necessarily have to complete a standards process in order to refer to it; that would put, for instance, the Zip format (used, among other places, in OOXML and ODF) out of scope for standards. WRT IPR issues: I think we just have to push forward on the assumption that all IPR holders who are part of the process will do their duty and disclose any relevant IPR, and hope that IPR held by nonparticipants in the process is not serious enough to cause us to regret our decision. Note: The statement about VP8 in the rtcweb-protocols document is a placeholder, put in there to indicate that we need to have the discussion. It's not a WG decision, but input to a WG discussion. Harald