ietf
[Top] [All Lists]

Re: S stands for Steering [Re: Should the IESG rule or not?]

2005-07-06 12:08:41
On Wed, Jul 06, 2005 at 05:24:40PM +0200, Brian E Carpenter wrote:
John C Klensin wrote:

--On Tuesday, 05 July, 2005 08:47 -0700 Bill Manning
<bmanning(_at_)karoshi(_dot_)com> wrote:


I don't believe that is true in this case, as long as RFC
2780 is in  force.
Especially since there is a clear path for Larry Roberts to
ask for IETF consensus, which by definition would overrule
the IESG.

  Brian



   just for my edification, how is IETF consensus determined,
   esp. when there is a need to "overrule the IESG"?  in other
   contexts, this type of thing is done by a vote of the
membership,
   but last i checked, there are no members of the IETF.


As far as I can tell, there are only two ways:

Middle posting: I think John and Bill are missing the
primary mechanism which is *trusting* the relevant AD to
call the IETF consensus honestly, even if s/he doesn't
agree with it. That shouldn't be hard to accept, since
we extend the same trust to every WG Chair.

If the AD makes the wrong call, then complaints, appeals and even
recalls are available.

        if John is correct, and your statements are correct,
        then IETF consensus kind of depends on IESG tacit
        approval...  and why would the IESG approve something
        that would "second-guess" or "overrule" a choice they
        already made?  Me thinks that we have a -VERY- high
        threshold to challange an IESG fiat.  And this may be 
        bad for the continued health/relevence of the IETF.
        Or not...  development can and does occur in other fora.
        The IETF may not see the work until it is cooked.
--bill


   Brian


    (i) We ask the IESG if they think they have been
    overruled.  In the nasty and impolite old days, the
    question was often stated at open plenaries, with at
    least the possibility of [real or virtual] over-ripe
    fruit moving through the air in the IESG's direction.
    Today, when a larger percentage of the community seems
    either inclined to suffer in silence or disinclined to
    speak up, the message from the community probably needs
    to be much more clear, and the consensus stronger, for
    the IESG to reach that conclusion.
    
    (ii) Someone appeals, essentially asking the IAB whether
    the IESG has been overruled.  The problem with this path
    is that, in the last decade, the IAB has tended to be
    deferential to the IESG's determinations about community
    consensus.  That determination is often hard to make,
    the various procedural documents appear to assign the
    job of making it to the IESG, and the IAB has tended to
    want to look at procedural issues on appeals and to not
    review the IESG's judgment calls.

Perhaps, in both cases, that is how we want it.  If it is not,
perhaps this is the time to start making procedural changes.
But, in either case, the difficulties implied by the above are
the greatest part of what makes statements such as the last
clause in the IESG's rejection of the Roberts registration
request particularly troubling: if the IESG has already
concluded the community consensus is sufficiently unlikely that
they recommend against trying to obtain it, the amount of
consensus, and ways of demonstrating it, that would be needed to
get them to decide their conclusion was incorrect and they had
been overruled may set an impossibly high barrier.

    john






_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
https://www1.ietf.org/mailman/listinfo/ietf