I'm not an expert on webrtc or codecs, but for what it is worth, I have some
opinions on the general issue of selecting technology in a WG. For full
disclosure, in my day job my colleagues have a stake in the matter, but I'm
just representing my own opinions.
Back to basic principles: Community decides. Interoperability. Running code.
Best technology.
Community decides: I think it is important that the WG and the IETF community
decides what to do in this case. The WG is in a deadlock, and they should be
able to decide - by consensus - if they want to attempt resolving this in a
particular way. Including alternate decision mechanisms. Of course, we all may
have opinions about how that resolution should happen. I'm personally in the
"coin toss" camp (but see further below).
And I'm reluctant for us in the management to attempt to override the decision
in any way. Lets learn what the community wants to do. If you do not like the
proposal on the table, make your opinion known or propose an alternative. Also,
I've heard a couple of arguments saying that we should be worried about appeals
on this matter. I'd say we should just try to the right thing.
However, the WG process is allowed to "fail" (if you can call it that) and not
standardise something that we do not find consensus on. We've been through this
many times in the past, and took various actions. Sometime it did lead to
considerable market split and lack of interoperability, sometimes a practical
unified reality emerged from what the vendors did, sometimes we learned to live
with multiple choices, sometimes we returned to the question sometime later and
found a standard. It has usually not been the end of the world.
Interoperability: This is one of the core values of the IETF. The reason why I
think it makes sense to try hard to come to a decision. Now, it may not be
possible, but having a common codec is very desirable. But in attempting to
come to a decision, it is important that we all try to understand the
real-world impact on interoperability. An agreement only on paper but not in
what products actually do is not so useful. Understanding the real willingness
of the parties to conform to what the RFC will say is important, no matter
whether we are talking about consensus, coin toss, or voting.
Running code: (Or running silicon, as it may be.) Another core value of the
IETF.
Best technology: Again, a core value. We should be careful about using
non-technical arguments in making a decision about technology standards. Of
course we all take all information into account. But technical arguments
usually have better longevity. Good trumps bad even tomorrow. Other arguments
are more susceptible to being changed, e.g., due to contracts, licenses,
passage of time, and behaviour of parties who might not even be involved in the
IETF today. Pretty much the only reliable information for the other factors is
past performance, which leads us back to running code and experience. But it is
indeed possible that on all objective criteria, two alternatives come out
equal. As an Internet user, in that situation I'd prefer the WG to make a
choice, *if* that choice actually leads the various players to be more unified
than us not making a choice.
Jari