Hi Stephen,
At 12:36 11-01-2013, Stephen Farrell wrote:
You mean rough consensus of the IETF I guess? Good question.
No, I mean consensus as that's also part what is gauged during a Last Call.
First, WG rough consensus is formally unaffected. As is IESG
review. And if IETF LC comments are received that do meet the
IESG discuss criteria then those should be handled as now.
I am not concerned about the WG rough consensus angle (for the
experiment) as it is a subset of a Last Call.
So I guess we're left with cases where there's a lack of
rough consensus during IETF LC but where the meat of the
disagreement is something that doesn't qualify for an IESG
discuss ballot.
Hmm, you mentioned "rough consensus" on an IETF Last Call. The IETF
Last Call is about "consensus". I may have misunderstood the
experiment as I did not read it as "rough consensus and running
code". To say it differently, there isn't any change to the IETF
Last Call; this is more about fast-tracking the WGLC and the IETF Last Call.
I'd say that'd be quite likely to allow the responsible AD
to say that there had been so much debate during IETF LC that
this experiment ought not be used.
Ok.
So, I don't think this experiment has any major effect
there really but maybe has a subtle one.
I'll be interested in seeing if this happens if we do the
experiment.
Let's see how the experiment works out.
I don't get what you mean. Can you explain? (I get that you don't want
to mention -bis cases, but I don't get why.)
What the experiment can do is make it easier to move from unpublished
specification to RFC. The "-bis" draft is from a previous RFC. If
you try to cover it the experiment can turn into a significant
process change. I prefer to leave that unspecified and see how the
experiment works out instead of telling people to use it for "-bis"
drafts. In a "-bis" document is in good shape it should not be a
significant problem to get it through the process.
Not a bad idea. Something like "fast-track must be started at least
one month (longer?) before an IETF meeting starts" ?
The fast-track process cannot be initiated within two weeks of an
IETF meeting.
You lose five weeks, two before, one for the meeting, and two after.
I guess the only problem I'd have with that is that for an experiment
that'd run for one year, that takes out about 4 months (3 meetings
and the year-end holidays) which is a lot.
Ok, make it one week then (see my previous comment).
If we extended the experiment to 18 months duration I'd have no
problem with something like that though. Or with leaving it as-is.
It can be left out as an implementation detail if it is easier to
keep the experiment to 12 months.
Can be. If a WG participant says "the decision you made on that
draft is broken: I appeal" then they first send that to the WG
chair.
Ok.
I don't see any text change being suggested there. Correct me if
I'm wrong.
You're not wrong.
Regards,
-sm