Re: DISCUSSION: rejecting candidates

[Donald Eastlake 3rd <dee3@torque.pothole.com>]

Hi,

On Fri, 12 Apr 2002, Brian E Carpenter wrote:

> Date: Fri, 12 Apr 2002 09:31:27 +0200
> From: Brian E Carpenter <brian@hursley.ibm.com>
> To: Thomas Narten <narten@us.ibm.com>
> Cc: Donald Eastlake 3rd <dee3@torque.pothole.com>,
>      ietf-nomcom <ietf-nomcom@lists.elistx.com>
> Subject: Re: DISCUSSION: rejecting candidates
>
> Ted's suggestion, I think, was to start one whole IETF sooner, i.e. pick the NomCom
> before the July/August meeting. Just moving up the end date alone doesn't work.
> Avoiding the December/January holiday season would certainly help, but we'd
> also have to avoid the July/August holiday season.
>
> However I have to register a continuing disagreement with Donald. From the
> viewpoint of the approving bodies (and I've served on both of them), there
> simply isn't enough time to go through an orderly candidate rejection
> process, and there is a clear sense that rejecting a candidate would be
> cataclysmic. If we don't fix this, the process doesn't actually have all
> the checks and balances we think it does.

I don't deny that giving a confirming body more time would allow them to
be more relaxed. I just said that (1) giving the nomcom more time was no
guarantee that it would send candidates to be confirmed for all slots to
the confirmation body any earlier, and (2) for most positions, rejecting
the candidate is not nearly as bad as you, and apparently the
confirmation bodies, seem to think because a 2nd choice is already more
or less lined up. (Admittedly, it is probably the case that for very
hard to fill positions, there is a higher probability that the candidate
is marginal or unusual and thus a higher probability of rejection and
their probably isn't a clear 2nd choice lined up and these are just the
positions for which the nomcom is likely to be using every last minute
to scrape up a willing and acceptable candidate.)

So why don't we just put in a paragraph stating explictly "The rejection
of a candidate by a confirming body isn't cataclysmic but is a normal
part of the procedss."?

The primary effect of organizational mechanisms isn't necessarily in
their execution. Because there exists a clear mechanism to remove Area
Directors, they act and are viewed by the community, to a greater or
lesser extent, in a different way than they would be if they held office
for a term of years or life (as members of the IAB used to before the
nomcom) without possibility of removal even if removal is sufficiently
rare that none has yet occured. Because confirmatory bodies exist and
candidates DO NOT take office until after they are actually confirmed,
nomcoms act and are viewed by the community differently even if
rejection is sufficiently rare that a case has not yet occured. Most of
the salutory effect of having a confirmation process has already occured
by the time the candidates are submitted.

The concept is not that the cofirmations bodies should be able to
leisurely take their pick of a variety of candidates. The nomcom is not
supposed to just be winnowing things down with the confirmatory body
making the final choice. As the representatives of the IETF Community,
the nomcom, with the advice and assistance of the liaisons, is supposed
to make the choice, subject to a final sanity check by the confirming
bodies.

See on additional comments below.

>     Brian
>
> Thomas Narten wrote:
> >
> > > While it all depends on the nomcom members and primarily on the chair,
> > > giving the nomcom more calendar time for its work will not always
> > > translate into it making decisions early.
> >
> > I agree. If you want to add some meaningful timetable wording to the
> > document...
> >
> > The current wording is:
> >
> >       The annual selection and confirmation process is expected to be
> >       completed one month prior to the friday of the week before the
> >       First IETF.  It is expected to begin 4 months prior to the Friday
> >       of the week before the First IETF.
> >
> > I think it would me more useful to say
> >
> > 1) the nomcom forwards its slate of candidates to the confirming body
> >    no later than (say) 6 weeks prior to the First IETF meeting. I.e.,
> >    allow sufficient time for the confirming body to think and maybe
> >    question.

Seems to me it would be useful to point out that it is normal for all
the IESG candidates to be confirmed by the old IAB before the new IAB
candidates are sent to the ISOC BoT. At least all nomcoms I know about
have done this because IESG positions are so much harder to fill and you
might want to filch someone from the IAB for them. So until the new IESG
is stable, you don't want to be pinned down for all the IAB slots. So,
in reality, the confirmations are serialized.

> > 2) The most important place for interacting with the community is the
> >    prior, i.e., Fall, IETF. To achieve this, add wording something to
> >    the effect: The nomcom should be fully constituted no later than
> >    one month prior the Fall IETF and should have made sufficient
> >    progress in its work so that it can use the Fall IETF to interview
> >    the community (i.e., IESG, IAB) as well as specific candidates for
> >    the open positions.
> >
> > I think it important to emphasize the point about "interviewing
> > specific candidates", in the sense that interviewing candidates
> > implies that the nomcom has actually started building enough of a
> > short list to identify some people it actually wants to meet with
> > face-to-face.
> >
> > Now, this may seem like its really moving up the target date. But
> > realistically, the December time frame where the nomcom needs to do
> > the most work, is really problematic due to holidays. This is both for
> > the nomcom members and the community (and candidates) when seeking
> > input/feedback.
> >
> > Another possible proposal: it might be useful to move the timing of
> > the nomcom to either Summer or Fall, i.e., so that the nomcom isn't
> > doing its work time over the December holidays.
> >
> > Thomas

Thanks,
Donald