Re: Proposed change to draft-ietf-nomcom-rfc2727bis-01.txt

[Dave Crocker <dcrocker@brandenburg.com>]

At 11:08 AM 7/12/2002 +0200, Brian E Carpenter wrote:
> > Let's take a step back, folks:  Why 5 months?  Why not 8 or 3?
>
> Specifically, because I think we need about 8 weeks more in the process,
> which I think is the intention of the draft anyawy.

Yes, it clearly is the intention.  What is NOT clear is how this extra time 
is actually to be used and why it is necessary.  Also...


> > 5 months means that we have a yearly process that takes nearly half a
> > year!  That just does not make sense.
>
> Why not?

What the heck, let's start the nomcom process 18 months before we select 
people.  That way we will be sure to have enough time.

And I hope that taking this ad absurdem highlights that we need concrete 
reasons for making these choices and that the reasons need to be 
balanced.  Besides that, spending 1/2 of every year doing selections keeps 
the tone of the IETF that much less stable.  It also makes affects that 
quality of the information acquired, since it permits that much less 
history for the people being evaluated.

Right now, it appears that the philosophy is to throw "time" at all nomcom 
concerns.  It therefore fails to consider what the actual problems are, and 
it fails to respect the time of the nomcom participants.


> >          It ignores the reality that a nomcom's members
> >          get exhausted.
>
> Surely having a bit more time to do the same amount of work should reduce
> the exhaustion level?

That is exactly wrong.

Nomcom is an extra activity.  The longer it goes, the longer the 
participants have to carry an extra activity.  The longer they have to 
carry the support of their management.  The longer the process of nomcom 
has to keep nomcom "project energy" strong.  The longer they are attending 
to nomcom and not IETF technical work or IETF management.

These are not abstract concerns.  They apply to any project and they were a 
very noticeable factor in the recent nomcom.


> > Sometimes, throwing money at a problem fixes it.  Same for throwing 
> time at it.
> >
> > Sometimes we need to figure out what real problems there are and fix them
> > more directly.
>
> And the real problem I spoke of in Minneapolis was the difficulty encountered
> by the conforming bodies in understanding and handling controversial choices
> made by NomCom.

I hope it's clear that this concern makes sense to me, too, and that it is 
a very good idea to attend to it.  However I believe that "time" is not the 
answer, but trying to improve the nature of the interactions -- and 
improving the sense of the role of the confirming body.


> > For all of the real and obvious concern over having the confirming bodies
> > feel rushed, I have yet to understand what would take place that would be
> > different and better.
>
> Having sat on the confirming side of the fence for 8 straight years, I can 
> tell
> you that feeling rushed when called on to approve a controversial decision
> is not a good place to be.

And just to be redundant, I'll stress that I agree that that is a bad state 
of affairs.  However it is not clear that giving the confirming bodies more 
time actually fixes the core problem, which I believe is to ensure that a 
confirming body has a solid basis for understanding how nomcom selections 
are made.

If that is not the core problem, or if there are more core problems, don't 
you think we should come to some understanding, before making assumptions 
of how to solve them?

For example, we could try to provide some information sooner.  And we could 
try to provide better information, as you have suggested.

(By the way, my sense of the activity in this nomcom working group is that 
it is dominated by IETF management and confirming body folks, and has very 
little participation by people who have been voting members of nomcom.)


> > Can someone please describe the changes to the confirming process that
> > would be enabled?
>
> There is no formal change. There is just time to do it with some thought.

And I repeat that that is not likely to improve things.  The 2 nomcoms I 
have been on both had push-back from one or both confirming bodies.  It was 
fine that the confirming bodies were concerned.

However the actual process was, essentially, that either a) the confirming 
body said the equivalent of "are you sure" and the nomcom said "yes", or b) 
the confirming body said "you need to talk to these people" and the nomcom 
did, with no change in the outcome.

Alternative a) is a waste of time.  Alternative b) has 2 problems.  One is 
that the confirming body did not adequately understand what interviewing 
had already taken place -- and frankly did not try to find out, though they 
thought they did.  (I think that your suggestion for proactive reporting of 
process details could substantially alter that.)  The other is that it 
basically represents an attempt by the confirming body to direct the 
details of the nomcom process.  It is my understanding that that is not a 
confirming body's job, and also there is no way a confirming body can do 
that at the end of the process, no matter how much time is allocated.


> > This treats the confirming bodies as collaborating members of the process,
> > which I suspect is the real change that needs to happen, as long as the
> > nature of the collaboration does not alter the role of confirming bodies as
> > confirming process.
>
> Yes, that wouldn't be bad, but perhaps as an oral tradition item until we
> have some running code.

sounds good.  so good that it should be applied to the desire to make 
calendar changes without having any real idea how they will be used...

\d/


----------
Dave Crocker  <mailto:dcrocker@brandenburg.com>
Brandenburg InternetWorking  <http://www.brandenburg.com>
tel +1.408.246.8253;  fax +1.408.850.1850