Re: [ietf-nomcom] BCP 10 Update, adding an IAOC Advisor to the Nominating Committee

[Spencer Dawkins at IETF <[e-mail]>]

Hi, John, and CC Alexey, who is shepherding this draft,

On Thu, Aug 24, 2017 at 9:58 AM, John C Klensin <[e-mail]> wrote:

>
>
> --On Tuesday, August 22, 2017 22:11 -0500 Spencer Dawkins at
> IETF <[e-mail]> wrote:
>
> > On your other points, I think I know what to do with your
> > feedback, but this one is worth talking about some more.
> >
> > There are different levels of "Nomcoms needing someone who
> > speaks IAOC-ese fairly fluently".
> >
> > I'm shooting for "don't forget to think about how you'll know
> > whether you've got a viable IAOC candidate to forward to the
> > confirming body, and if you don't know who can help, the IAOC
> > should be well-placed to make suggestions about people
> > who can help".
> >
> > I could be shooting for "the Nomcom has to ask for help", or
> > even "has to ask the IAOC for help".
> >
> > At the extreme, I could be shooting for "change the definition
> > of committee membership so that if you don't have
> > representation from the IAOC, you've got a really big
> > problem", to match not having a liaison from the IAB or IESG.
> >
> > Are people comfortable with this being more permissive than
> > prescriptive?
>
> Spencer,
>
> I'm almost always comfortable with more permissive or, more
> specifically, with guidelines and principles rather than rules.
> That is in part because I think the IETF has a terrible track
> record when we try to make specific rules, almost always
> discovering either that they are not quite right and require
> either workarounds or more effort to try to fix or because they
> lead to quibbling about the rules rather than getting useful
> work done.
>
> However, as I've tried to say before, I'm worried about a
> different problem in this case, which is whether, as the
> individual knowledge of Nomcom members goes down, the presence
> of all of those liaisons could have undue influence by any of:
>
> (i) influencing the Nomcom for or against particular candidates.
>
> (ii) influencing the Nomcom too much about the role of the
> relevant bodies or job descriptions.  That is especially
> hazardous because part of the original intent of the Nomcom was
> that it be able to make decisions that particular bodies were
> going in the wrong direction and then make appointments with the
> express intention of changing that.
>
> (iii)  Having a chilling effect on community members who wanted
> to make comments (favorable or unfavorable) about particular
> candidates that they wouldn't want generally known, or known to
> the candidates or their supporters or opponents... especially
> when the liaison was believed to have a strong (again, whether
> positive or negative) relationship with that candidate.
>
> These are obviously far more general issues than how to add an
> IAOC liaison or how, but I believe that specifying an additional
> liaison, especially from a body with the composition and role of
> the IAOC, could considerable amplify any issues or concerns of
> that type.
>

I have seen your e-mail about that, and I think I understand this issue
pretty well.

I agree that how the Nomcom reviews positions and nominees has changed
pretty massively since 1992.

The problem with problems with Nomcom is, it's really hard to know what
problems have been encountered over the years, unless one or more Nomcom
chairs reveals a problem that they encountered.

Some do that, pretty publicly, but even in those cases, people outside the
Nomcom don't have the details that would be helpful in solving problems.

I might reasonably point out that there's been an "IAOC liaison" on every
Nomcom from 2009-2010 to 2014-2015, but maybe we've just been lucky - I
just don't know.

So, here's what I'm thinking.

Tl;dr

Alissa (as Gen-AD) is out on leave until October (although she did surface
for an informal telechat earlier today).

I'm not comfortable proposing more than minor changes, until she can
participate in that discussion.

When she returns, I'll tell her what I'm hearing (and that I can't dismiss
that concern out of hand). That's on my calendar as a reminder.

The long part ...

The last time we made a Nomcom process change that was more than a
paragraph long, what Russ did, was convene about ten years worth of Nomcom
past chairs to exchange, maintaining as much confidentiality as they could
(so, no names), descriptions of problems they'd encountered. Common
problems went into
https://tools.ietf.org/html/draft-dawkins-nomcom-3777-issues-00, and that
formed the basis of most of the BCP changes I authored (most notoriously,
https://datatracker.ietf.org/doc/rfc5680/). That's why we had some
confidence that we were talking about real problems.

ISTM like it's worth doing something like that, every decade or so, which
is how long it's been since I was working on that draft.

Maybe all the problems that were identified then have gone away without BCP
changes, or maybe more recent Nomcom chairs have encountered different
problems that are worth thinking about, but (inserting by reference every
conversation you and I have ever had about how hard it is to get BCPs right
the first time), if we've already had problems that recur often enough to
justify BCP updates, I'd like to understand them.

Thanks,

Spencer


> best,
>      john
>
>
>
>