Re: Possible extra "oral tradition" for draft-ietf-nomcom-rfc2727bis-01.txt

[John C Klensin <john+nomcom@jck.com>]

--On Friday, 12 July, 2002 05:39 -0400 Fred Baker
<fred@cisco.com> wrote:

>...
> Similarly, in looking at the IETF Chair (which the nomcom
> appoints) or the IAB Chair (which the nomcom does not appoint,
> but the question of who will likely be chosen to do the job
> does come up in discussion), we are talking about the CEO or
> CTO of the organization in some sense. What are the
> qualifications of a CxO? I suspect that they can be difficult
> to enumerate precisely, but come down to the factors one
> considers in answering the questions
>...

Fred,

I had hoped to sit this particular conversation out, given
concerns that someone might worry about my motivations.  But I
think you raise, in passing, an important point, so I will risk
it with the understanding (in the event that it was not clear
from my remarks in Minneapolis) that nothing could ever induce me
to take the IAB Chair position again...

The notion of the nomcom trying, as part of its decision-making
process, to second-guess various decisions over which it, in
theory, has no control is troubling and, IMO, a threat to the
system.  The IESG appoints a liaison to the IAB, various ADs de
facto appoint liaisons and representatives to other organizations
that fall within the scope of their areas (the IAB must approve
these, but, absent an obvious problem, the approval is pro
forma).  The IESG determines what areas exist and how many ADs
are allocated to each one.  The IAB appoints its own chair, an
IESG liaison, liaisons to other organizations (which, under
evolving scenarios, includes ISOC Board seats and ICANN
representatives).  The IAB Chair appoints the IAB Executive
Director.  The IAB and IESG Chairs, jointly, have effective
management responsibility for the IETF Secretariat, and the IETF
Chair has ultimate sole responsibility for where meetings are
held.

I think that the Nomcom should look as broadly as possible at the
qualifications and likely responsibility of the people it is to
select.  But this predictive collateral stuff can, I think, get
us into too much trouble by providing essentially bogus reasons
for making (or not making) a selection.  Think about where it
leads:  

	"She would be a really good AD, but the IESG might make
	her liaison to the IAB, and we don't want her there".  
	
	"He would be a good IAB member, but he is a bit too
	interested in ICANN and we don't want any chance that the
	IAB would put him on the PSO-PC".

	"If placed on the IAB, she might end up Chair, and that
	might give her too much influence, so let's keep her out".

	"I really think we need to have a separate IPv6 area, and
	he is opposed to that, so let's keep him off the IESG
	even though he would be a good AD."

	"She wouldn't otherwise be a very good IETF Chair, but my
	first priority is to get an IETF meeting held in Pago
	Pago, and she has promised that would be a priority, so
	I'm going to push for her."

I suggest that these are gray areas (I assume that the last one
would clearly be rejected by the rest of the Nomcom as bogus, at
least if the reasoning could be identified, but the others may be
a bit more subtle).  Assuming the Nomcom is being responsible,
they appear to also put too much pressure on, and decision-making
power in the hands of, the respective liaisons: making such
decisions rationally requires predicting how the IESG or IAB will
behave internally, and the liaisons are the only
presumably-reliable sources of information on that topic.  

To make this more complicated, we have no guidelines on how the
liaisons are expected to behave vis-a-vis the bodies that
appointed them: in my tenure on the two bodies, I saw liaisons
who would actively consult those bodies (or the non-candidate
subset of them) when a question came up in the Nomcom about their
operations and saw others who interpreted the confidentiality
rules as preventing them from doing so (sometimes, I assume,
reinforced by personal preferences or style).

The Nomcom needs to trust the people it nominates to do those
jobs competently and responsibly -- as the IETF Chair, as ADs, as
IAB members.  If it does not, it has no business nominating them
for anything.  But, for these secondary decisions that are
ultimately the result of how the IESG or IAB decide to organize
themselves -- responsibilities the current procedures fairly
clearly give to the IESG and IAB -- the Nomcom should, IMO, avoid
guessing and, ideally, be barred from it.   If that isn't what
the community wants, then we need to change the organization
structure, not encourage the Nomcom to tamper.   For example:

	* There are arguments for the IESG deciding about area
	organization and allocation of ADs to areas (e.g., only
	the IESG can really figure out what it needs).  There are
	some arguments the other way (e.g., the community may
	have a better idea).  At present, the IESG decides.  We
	could examine the balance, decide it gives the IESG too
	much power, and change it.  But, if we don't, the Nomcom
	should avoid trying to figure out how, by manipulating
	appointments, it can change the IESG and Area structure
	(and a Nomcom member with that agenda should be, IMO,
	replaced).
	
	* There are arguments for the IAB picking its own Chair
	(e.g., the IAB needs to function as a group to a greater
	extent than the IESG and having it select its own Chair
	is its best protection against a poor choice).  And there
	are arguments against it (e.g., including  the perception
	that the Chair has too much external or individual
	influence).   Again, at present, the rules are such that
	the IAB makes that appointment from within its own
	membership.  If that is the right rule, the Nomcom
	should, IMO, stay out of it.  If it is the wrong one,
	then we should fix it, either to make the IAB Chair a
	Nomcom appointment like the IETF Chair or, in principle,
	to remove the restriction that the IAB Chair must come
	from the sitting IAB membership.  But the predictive
	second-guessing strikes me as unwise and, ultimately, as
	an attempt by Nomcom members to circumvent fairly clear
	procedures about scope of Nomcom responsibilities.

	* There are arguments for the IAB selecting liaisons and
	for extending that responsibility into the PSO-PC, the
	ISOC BoT, potentially the ICANN Board.  There are
	arguments that some of these have impact too sweeping and
	that they should be entrusted only to the Nomcom or to
	some process yet to be defined.  But, again, if the
	community doesn't want the IAB making these appointments
	(or even deciding whether they should be made), it needs
	to engage on the issues and decide who or what should
	(rather than ending up in the rather inconclusive
	discussions on the Poisson list about one of these topics
	a year or two ago).  But the Nomcom is just not selected
	to make those decisions for the community by attempting
	to manipulate the composition of the IAB --these are not
	decisions that should be made in secret, nor by a body
	that doesn't ever have to explain the reasons for the
	decisions it is making.

Just my thoughts, but ones based on a certain amount of
experience.

       john