Re: PROPOSED GUIDELINE: balance

[Theodore Tso <tytso@mit.edu>]

On Thu, Apr 04, 2002 at 05:56:41PM +0700, Robert Elz wrote:
> One think that should be made clear is that the nomcom can ask
> questions (especially where they don't involve specific nominees)
> of anyone they like.
> 
> If the nomcom had a question about what the makeup of the IAB should
> be, they could ask the IAB, the IESG, the ISOC, or the whole IETF
> (via the list, or if the question arose before the December IETF
> meeting, at the plenary).

International representation was certainly mentioned as something
important to consider, but not necessarily as being more important
than other issues.

I have heard at least a few people comment that this year's IAB slate
resulted in a "dramatic step backwards" in terms of international
representation on the IAB, and that this might hurt the IAB's
credibility in certain regions of the world, such as Europe and Asia.
The extreme version of this complaint is that in order to improve
IAB's credibility when dealing with I18N and ISO/ECMA issues, Nomcom
should use an Quota-based, Affirmative Action plan to select at least
some number of non-North American (preferably a non-native English
speaker and preferably not residing in an English-speaking country)
candidates on the IAB.

(n.b.  I doubt most of the people who were unhappy about the resulting
international representation would be happy with a quota based
approach; I just used an extreme option to make a point.)

Some people might think that paying attention to the overall
international representation on a body is naturally part of RFC
2727's, "superior candidate" wording, but I'm not convinced; it's so
vague that it can mean anything, which means it means nothing.  

So if there is ever a question about picking the "best" technical
candidate from the U.S., and someone who is "not-quite-as-good, but
still quite decent" candidate from a non-North American country, what
should the nomcom do?  Different people will have different ideas of
what constitutes "superior".  I'd much rather have this discussion
publically, even if we can only have it at an abstract level, instead
of having it closed doors.

> I'm not sure that embedding any particular guidelines in the nomcom doc
> is a good thing - the desirable attributes tend to change over time,
> which is very easy to handle in e-mail to the nomcom but much harder if
> there is some suggestion buried in an RFC.   If the nomcom was faced
> with choosing either someone who provided an attribute suggested by
> the nomcom RFC, or choosing one who satisfied some requirement communicated
> to it by the current IAB, but couldn't satisfy both, which should it
> prefer?    My guess is that the RFC being the "will of the IETF" would
> win, it would have to - but that's almost certainly not the best outcome.

I suppose the question is how frequently do we think the desireable
attributes change over time?  If we think they change very frequently,
then I could imagine a process where every year, there is a 2 month
window where there is an open discussion mailing list where selection
criteria in the abstract (no names, or actual candidates would be
discussed)for the IAB and IESG could be aired.  This could serve as
additional input to the nomcom.

If we think they change every 3-7 years, then I'd argue that it should
be codified in an RFC, and we should simply remember to revisit and
re-update the RFC every couple of years.

However we do it, I very strongly believe that the selection criteria
is something that should be discussed publically, and not left in a
secret, smoke-filled room.  Let's only keep secret those things which
really need to be kept secret.

						- Ted

P.S.  Actually, with respect to the international representation
issue, a much more moderate position is that Nomcom should spend a lot
of time searching out and arm-twisting non-North American candidates
and getting them to agree to be consider --- since there are qualified
people out there, it just might take more effort to find them.  One of
the other benefits of having a more stretched-out timescale is that we
would have more time to find such folks.  In the current time-scale,
things are really very rushed, and it makes it a lot harder to spend
time focusing on niceties such as spending time to do arm-twisting
just because there's "not enough candidates from the Asia-Pacific Rim"
or "not enough candidates from Europe".