Longer or more meetings?

[Henning Schulzrinne]

Other SDOs, such as W3C, rely much more heavily on teleconferences, 
typically weekly. With modern web-based meeting support tools (shared 
whiteboard, shared slide viewing, text chat for questions, etc.), these 
can be fairly effective. They also avoid the "selection by travel 
budget" problem that you do not mention below.

One should also note that other SDOs typically have schedules and 
deadlines that mean something. It doesn't mean that things get approved 
when the buzzer sounds, but it does mean that the WG is expected to 
produce results on a definite day. It's apparently not uncommon that, 
say, IEEE output gets voted down or has to go through a second 
iteration, but at least it doesn't drag on forever.

In the SIP* working groups, the externally imposed hard deadline 
(courtesy of 3GPP) had, in the balance, a healthy effect. The document 
got produced at the 98% quality level, and it probably got produced one 
year earlier than it otherwise would have been. Among other things, it 
forced an efficient discussion style - file open issue, discuss, close, 
next. No interim meetings were needed.

At least in W3C, keeping things on schedule is a lot easier since each 
WG seems to work on one (master) document. I think that's true for IEEE 
as well. It's a lot easier to see progress, or lack thereof, on one 
document than on half a dozen or more, where discussion on the mailing 
list flits back and forth at random and where each topic gets five or 
ten minutes of discussion/presentation time three times a year.

John C Klensin wrote:
> (this is substantive, so I'm starting a new thread)
> 
> Based on Randy's comments, Jari's, and a few others, let me suggest 
> another strawman.  Note that I'm not at all sure this is a good idea, 
> but it probably belongs on the table.
> 
> Problem: Face time of a few hours three times a year is not sufficient 
> to make progress.  Many WGs discover that they need interim meetings to 
> really move forward.
> 
> Suggestion: Insist that WGs conduct interim meetings so they could face 
> off in day-sized units, not slots of an hour or two duration.  At area 
> option, cluster these interim meetings into several-day sequences, with 
> overlaps and sequences to be determined by the ADs.  Cut full IETF 
> meetings back to once a year, and focus their schedule around 
> interactions, cross-fertilization, and cross-checking within and between 
> areas and on issues of IETF-wide importance.
> 
> Possible downsides: This might push us toward professional 
> standardizers.  It might reduce the quantity and quality of input from 
> folks with wide perspective.  It might effectively kill the notion that 
> we do most of our work, and all of our consensus-determination, on 
> mailing lists rather than in meetings.  We might easily slip into a 
> model in which WG meetings were attended by working engineers, but that 
> IETF meetings were attended only by professional standardizers, "process 
> experts", and general-purpose go-ers.  It should be noted that some of 
> the standards bodies whom we often deprecate use a model roughly similar 
> to this and have fallen into those patterns.
> 
> If we need broad coordination more than annually, e.g., if we have WGs 
> that need inter-area examination and interactions but that are likely to 
> complete their work in 12-18 months, this could be a real quality disaster.
> 
> Possible useful side-effect: if we made it clear that we expect ADs to 
> get to most of the interim meetings of their WGs, it would provide a 
> powerful incentive to them to keep the number of WGs in their areas 
> small and of brief duration.
> 
> Is this worth thinking about?
> 
>     john