John,
On Tue, Sep 20, 2005 at 09:37:43PM -0400, John C Klensin wrote:
If it is possible, I'd be most happy to see this viewed as an
experiment in the general spirit of RFC3933 but without some of
the trappings. Treat it as a real area, appoint AD(s) as
appropriate, but go into it with an agreement and the
expectation that it, and the area structure more generally, will
be carefully reviewed, and these tradeoffs reexamined, in a year
or so. If it turns out to cause more problems then it is worth,
perhaps we could then get rid of it --with no assumption of poor
behavior on the part of the ADs-- and try something else.
I am all for experiments but this one would be particularly tricky to
setup. It is extremely hard to measure whether productivity of the
IESG increased in a given period as none of the determining factors
remain unchanged. Eg., the number of appeals in a given period is not
the constant, the number of documents that we process is not the same,
the number of new working groups considered is not the constant, the
total number of working groups that is being managed is not constant
either and finally the cast of area directors (outside of the area
directors that would be added) is different too.
I wonder how productive it is to do an experiment about efficient
group size while we already know that it hasn't worked in so many
other settings. It is not like there are no warning signs already that
we are spending already a significant amount of time on overhead
instead of doing the work that we were appointed for.
At some point we have to decide that the IESG cannot become any
larger. It is very easy to add one/two ADs at a time. Each time, it is
unlikely that the addition of a single AD would cause the system to
collapse. However, at the same time this is a very slippery slope. I
believe that at some point we have to draw a line in the sand.
Considering that we are already larger than optimal, now would be good
time to make such a decision.
David Kessens
---
_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
https://www1.ietf.org/mailman/listinfo/ietf