[Ietf-hub-boston] The October health of the IETF discussion (was: Re: Future meeting schedule)

[John C Klensin <[e-mail]>]

Dave,

I started that discussion because I'm very concerned about where
the IETF is headed.  As Kyle said, not to gossip about (or
trash) individuals but because I see some medium and long term
trends that we, as a community, are not handling very well and
may be in denial about.  Part of my concern is that, if those
trends and our [non-]response or arguably inappropriate
responses continue, we could easily find the IETF irrelevant or
worse.  In a way, the leadership is probably a symptom rather
than a cause although, if one wanted to point fingers, some
people have made things better (or slowed the speed at which
problems accumulate), some have made things worse (or
accelerated those trends), and some people have been in one
group on one occasion and a different one in another.

My plan had been to ask permission to record the discussion so
it could be made available later if it turned out to be helpful
(or at least eye-opening or informative) and I brought my trusty
voice recorder with me, but then I spaced it out and didn't
realized I had done so until I was in the elevator on the way
home.  Sorry about that.

Others would probably identify other key themes --it was an
interesting discussion that went in ways I didn't anticipate--
but one was the one Kyle mentioned.  From my point of view, part
of what drove POISED and the reactions to the perception that
the IAB had gotten seriously out of touch with the community was
a desire to identify and make the most obvious and important
changes quickly, get perhaps 80% of the perceived problem(s)
solved, and then more on.  The result of working that way are
almost exactly the same as they would be if we approached
network engineering the same way: one deploys the solution,
things mostly work, but, over time, the number of loose ends
that such a process inevitably leaves come back and cause harm,
whether in the form of insidious bugs, failures of robustness,
inability to adapt to evolution and other changes, and so on.  I
think it is only one important example, but an accidental
consequence of the old model of an IAB in control and spawning
and overseeing various task forces is that the IESG's role was
strictly to manage and steer the IETF.   ADs didn't find
themselves responsible for chartering, overseeing, nurturing,
and generally supporting a WG and its work, only to then have to
turn around and try to support an impartial and broader-scope
evaluation of that work that might result in its rejection.  We
changed that with POISED and the reassignment of final document
review to the IESG, creating that conflict.   All in all, I
think it has been managed astonishingly well (far better than we
had any right to expect) over the years, but, as things get more
specialized (maybe an inevitable result of our working on a 40
or 50 year old technology and its derivatives rather than at a
much earlier stage but YMMD), I think we are seeing more
frequent symptoms of that change being more of a "best thing we
can think of quickly" decision than one that was carefully
thought out with the whole system and possible side-effects
considered.  

I was at a memorial service for Corby Monday afternoon and I
think it was Jack Dennis who made the observation that, while
Computer Science turned into a real academic discipline during
Corby's career, we still haven't gotten there yet with Systems
Engineering and that is a problem.   In a way, that connected to
the concerns about and the IETF hasn't gotten there, or stayed
there for long, either.

Anyway, the discussion was interesting and, at least to me,
helpful, informative, and helped me calibrate some of my
thinking on the subject.  Sorry you missed it.

best,
   john


--On Monday, November 4, 2019 15:05 -0500 Dave Lawrence
<[e-mail]> wrote:

> Kyle Rose writes:
>> Lest this go off the rails, let me be clear: by "talking
>> about the IETF leadership" I don't mean we spent the time
>> gossiping about  individuals. 
> 
> Well that's good.  I hadn't figured it was that way, though I
> can see how some might have taken the "eye-opening" remark in
> that vein.
> 
>> It was more about how the very structure of the leadership
>> (for instance, the modern relationship of the IESG to the
>> IAB) influences decision-making, outcomes, and even (in my
>> interpretation at least) the Overton window of
>> publicly-acceptable viewpoints.
> 
> I'm still really interested in this.
> 
> _______________________________________________
> Ietf-hub-boston mailing list
> [e-mail]
> https://www.ietf.org/mailman/listinfo/ietf-hub-boston