what are the real problems

[John C Klensin]

--On Friday, 23 May, 2003 11:25 -0700 Dave Crocker 
<dhc at dcrocker.net> wrote:

> Competing efforts have a natural filter, called critical mass.
> If they can't develop a critical mass of serious effort, they
> die of their own accord. If they *can* develop competent
> effort, then it is essentially hubris for IETF management to
> arbitrarily choose between competing proposals.

Dave,

While I think I agree in general, I think we need to be a bit 
careful here.  It can sometimes be exceptionally difficult to 
tell a mob from "critical mass" and even more difficult to 
distinguish between a pack of wolves in full chase and "strong 
consensus".  Both mobs and wolf packs typically have some 
organizing leadership, although often, when the leaders are good 
at their work, the only clear symptom is that everyone is headed 
in the same direction, or baying more or less in unison, rather 
than running off in all directions at random.

If something dies from lack of critical mass, fine: we generally 
should wish the effort well and speed it on its way to a happy 
rest.  I'm really pleased about the number of WGs the IESG has 
succeeded in closing in the last several months --both the 
"closed/sucessful" and "closed/wasn't doing much" ones-- I think 
it is a great trend that should be applauded and continued.  But 
the combination of "they have critical mass" and "there is 
competent effort" are wildly subjective and both are subject to 
gaming.  And what is, and is not, "arbitrary" is even worse: I 
don't think any of us would argue for, or encourage, the "IETF 
management" to make arbitrary choices between competing 
proposals (or any other arbitrary choice unless a choice is 
important and the particular choice really makes no difference). 
But that doesn't imply that some work with WG consensus behind 
it shouldn't be killed, or that choices should not be made on 
the grounds that one is better for the Internet than the other, 
and so on, even if the WG doesn't agree with the reasoning.

I'm not opposed to competing standards, although I'd hate to see 
us make a habit of it (and, in most cases, would prefer 
competing Experimental protocols with a clear plan about review 
and real decision-making).  I think that there are some 
circumstances in which we really should say "nice try, but no 
sale and no standard" to WGs -- again, I would be very upset if 
it became a habit and extremely concerned if it emerges as a 
"late surprise", but I think it is dangerous when the IESG 
doesn't believe it has that power and, indeed, responsibility to 
make those decisions when the WG has been getting, and 
resisting, push-back for some time without any real effort to 
engage in a dialogue on the issues the AD is raising.

We have both seen situations in which a relatively 
well-organized, but very narrowly-focused, group comes to the 
IETF wanting either ratification, or an umbrella organization 
for, standards it wants to produce.  These groups have critical 
mass by any objective measure we could apply. What they often 
lack is a strong sense of responsibility to and for the Internet 
as a whole as distinct from their narrow efforts and focus. 
Sometimes they can be educated, sometimes they can't.  But they 
routinely argue "critical mass" (often in the form of how many 
people they can get to hum loudly at a BOF) as evidence for 
their entitlement to have a WG.  If they get such a WG, it often 
requires superhuman efforts on the part of the relevant AD to 
get and keep their work on-track with basic functionality and 
interoperability goals on the public Internet, and the ADs are 
often damned for those efforts when we, as a community, should 
be thanking them.

The same situation applies late in the process: if there is a 
controversial issue, and dollars are at stake, it is rarely 
difficult for someone who is good at whipping up crowds to get a 
number of folks to yell loudly in unison either for or against a 
proposal or some push-back about it.

Ultimately, someone needs to make a judgment about whether a 
particular size group, or a particular volume of comments 
--favorable or unfavorable-- represent true "critical mass" or 
"rough consensus", and whether the result is competent, or 
whether they represent a possibly-well-organized effort to push 
a special interest or insufficiently-thought-out protocol 
through the IETF.  Even from my perspective as someone who has 
had more than the usual quota of disagreements with the IESG as 
a whole and with how individual IESG members have handled 
particular situations, I think our best hope is to let --and 
expect-- the IESG to do its job in these areas.  That means 
avoiding getting into the kind of second-guessing in which "John 
(or Dave) thinks there is critical mass" or "everyone who spoke 
more than twice at the last plenary agreed" are presumed to take 
precedence over IESG deliberation and decision-making.

Does "let the IESG do it" have a potential for abuse?  Sure.  Do 
I, personally, think it has been abused occasionally?  Yes.  Do 
I believe they have made the wrong decisions, or have been much 
too slow to make a decision, sometimes?  Yes, more certainly 
than I think there has been intentional abuse in getting there. 
But, for those issues, I think we need to concentrate on having 
the basis on which decisions are made be much more public and 
transparent.  I think we also need to look at improving the 
efficiency and effectiveness of the checks and balances 
--including the appeals and recall procedures-- if we conclude 
that they aren't usable enough.  But I think that blanket 
statements about criteria that, themselves, are subject to abuse 
and require subjective interpretation just take us around in 
circles.

> For all of the times we have efforts to create competing
> proposals, most die of their own lack of credible technical
> support. They do not need arbitrary management 'help'. For the
> few that are serious, we do not need to worry about the
> "drain" on IETF management.

In general, yes.  But, perhaps because I seem to like worrying, 
I don't want to do anything that would say to the IESG "even if 
you think this is important and should be encouraged, if they 
don't have critical mass on their own, you are not to go try to 
nurture them or figure out ways to move the effort along".  And 
I do think it is important to examine, or help the IESG examine, 
their workload so that, if something _is_ serious, the cycles 
are available to deal with it without setting off crises about 
how other things are getting delayed or "blocked".

regards,
    john