Is this really where we want to go? (was: Re: Selecting leadership, take 2)

John C Klensin john-ietf@jck.com
Sat, 07 Dec 2002 11:20:24 -0500


--On Saturday, 07 December, 2002 16:31 +0100 Harald Tveit 
Alvestrand <harald@alvestrand.no> wrote:

>>> [ just to highlight your last paragraph ]
>>>
>>> maybe shortness of meetings actually is a problem.\
>>
>> That is why I tried to make the specific example extreme
>> enough to be silly.  The topic deserves, IMO, discussion.
>> And I find it interesting, and useful, that you did exactly
>> what I was advocating (and that Ted followed up on in more
>> detail and more eloquently than my original statement): to
>> oversimplify, you stated the possible problem ("shortness of
>> meetings"), explained the presumed bad effects ("slow
>> convergence and poor quality")
>
> actually I would claim the opposite nomenclature here - that
> the "slow convergence and poor quality" is the problem, and
> "shortness of meetings" is one possible cause (other possible
> causes being lack of ability to force convergence on mailing
> lists, entrenched company positions, lack of committment of
> resources from the people who need to work on the problem,
> lack of early architectural input...)
>
> but this is, perhaps, syntax, not semantics.

I think so.  But, personally, even that is a more interesting, 
and potentially fruitful, question than "can we construct a 
comprehensive problem model".

>> , identified a remedy other than the four-week IETFs of my
>> strawman (interim meetings), and identified a possible
>> downside ("standards professionals" and "loss of wide
>> perspective").  So, I think, we should now be able to have
>> all of those discussions (and I am pleased that two have
>> already started), including, explicitly, the "email versus
>> meeting time" tradeoff that Jari raised.
>>
>> That, it seems to me, is how we make progress.  And, while the
>> implications are broad, the   The alternative of trying to
>> build a comprehensive problem model feels, instead, like a
>> way of going around in circles.
>
> possibly my problem is that I don't feel that we've made it
> around the circle even once yet....

That is fair.   However, from my point of view, a case can be 
made --and, in different ways and focus, has been made by 
several people-- that circle-traversal is neither a necessary 
nor a particularly desirable approach.  We may be wrong, but I 
haven't heard and understood a clear argument as to why (that 
may be my problem, of course).  But, if we don't need to 
practice circle-navigation to make progress, I don't see much 
reason to go around even once to _prove_, experimentally, that 
it is unnecessary and pointless, or that circles are connected, 
or whatever it would demonstrate.

So I'm listening, anxiously, for someone to make the case

	* _for_ comprehensive problem-model building or
	
	* why trying to head directly into workng out solutions
	for specific issues we can identify is undesirable

For example, one could argue that everything we do, and all of 
our processes and rituals, are sufficiently interconnected that 
it is risky to do anything without understanding the complete 
picture.  I think I would reply by asking why, were that true, 
the IESG has been making as much policy as it has in the last 
few years under the labels of "nits" or operational changes. 
But no one has made that case at all, at least in my 
recollection -- the concept was just assumed and people seemed 
to buy into it by, e.g., debating qualification for WG chairs 
under that scenario.

So, please make it.  Or explain why going around the circle at 
least once is desirable.  Or, let's try something that might, at 
minimum, be reasonably expected to go faster.

regards,
   john