On 12/06/2013 03:50 PM, Eliot Lear wrote:
Gonzalo,
Thanks for this really well written and good summary. Again, the chairs
– and the area directors – as well as the whole RTCWeb group have been
working really hard on what many of us think will be critical
technology. As one of the whiners about voting, let me say first that
*do* believe that strictly speaking the rules can be interpreted to
allow voting. That doesn't make it a wise choice or a *substitute* for
rough consensus. Therefore I subscribe to the following interpretation:
On 12/6/13 10:57 AM, Gonzalo Camarillo wrote:
Other people think that using
processes such as straw polls or some types of voting can help the WG
understand better the situation at hand and help build consensus, which
will be *ultimately* evaluated in the WGLC and IETF LCs on the document.
And I would go further, and suggest that RFC 2418 provides working group
chairs with very broad latitude to declare what rough consensus is in
such circumstances. The document merely points out that it is > 51% and
it needn't be 99%. That large range should be fair warning that the
ruff in this case could be quite rough. The ultimate point, as Jari
mentioned in his note of 2 Dec 2013 is to aim for broad interoperability.
And again, as one of the whiners, I probably have a responsibility to
put forward some thoughts on how one might proceed. I state this as an
outsider to the WG, and so perhaps these ideas have been already tried
or considered and discarded.
You've created a rather long list of options. It seems to me that it
would be good to know what we are talking about in terms of "rough". It
might be useful still for participants to be polled on whether or not
they could live with each of those options. That is - have a positive
signal. This can be used to establish a front runner or group of
popular choices. Recalling Jari's message again that interoperability
is what we are aiming for, it might also inform the working group to
understand what user bases and implementations are impacted, at the same
time. Stating which implementation one is developing or is planning to
deploy will hopefully provide some feel for interoperability. This does
weight near term over long term, but I think that's a good approach (you
don't get to long term if you can't satisfy near term).
This is one aspect I like about the "vote" (distasteful as I find the term).
A full stack-ranking by each participant of the alternatives on the
table gives
a clear idea of how they weigh different alternatives against each other
- which
allows us to probe more deeply into what possible compromises might actually
be possible for participants to live with.
The chairs have promised that the actual votes (full ranking + names) will
be made public. That allows everyone to analyze the stated opinions and see
what they interpret from that information; I think that's a Good Thing.
I offer these purely as suggestions to be discarded if they are not
found to be useful.
Again, I wish the WG and the chairs the very best in trying to untie
this knot, and I offer to be whatever use I can to assist.
Eliot