MAJOR ISSUE: WG formation process

[Allison Mankin]

> 
> Having looked through the archives, I haven't seen much discussion on this
> topic, so I thought I'd bring it up. Is there a "problem" here?
> 
> I'd be interested if anyone has looked at the time it takes to get a WG
> going -- that is, the time between when the first BOF occurs and when the
> announcement of WG chartering occurs.  If our focus is timeliness and
> relevance, then this also needs to occur in a timely way.  In general, I'm
> curious as to the minimum, average, and maximum times that is required for
> WG formation and how this compares to say, the time it takes to issue the
> first document once the WG is chartered.
> 
> Of course, comments on other aspects of the WG formation process -
> predictability, transparency -- are also welcome.
> 

Bernard,  

There are aspects of relevance, IETF-wide overload, concentration of
power, and transparency, all important to the WG formation process.
I thought I'd send a summary note from my perspective of having helped quite
a few WGs into existence.  It takes the form of a chronology of the process
when there is just one BOF.  Just getting the process clear and transparent
is something we have never done, and which "problem" should do. There are 
issues that I can see.  More comments specific to "problem" are interspersed 
at **.  

BOF request
  o Does the problem belong in IETF?   BOF may be used to test.
      Area's "charter", big picture, constituents, state of industry, 
      judgement calls, reviews, all enter in...this part is not Process,
      but our real intellectual capital, our stock in trade...I think the
      process here applies to ensuring that the broad enough eview of 
      whether it's an IETF problem takes place.  More below on that.**
  o Is there a good charter?  AD may iterate with proposers, with 
    discussion on BOF mailing list.
  o Is there a draft clearly showing the problem?  Discussion begun?
      The Transport Area does not allow BOFs without.
  o Even if the problem does belong in IETF, are there cycles
    to do the proposed work (management and development)?
      Can the proposer show before or during the BOF that workers
      will come forward to justify IETF investment in the effort?

BOF review
  o IESG/IAB BOF pre-review.  Process needs more transparency**.
  o How did the BOF go?  Meeting, minutes, mailing list...
  o IAB review.  Process needs more transparency**.
  o Determination of outcome from meeting consensus and reviews.
       This determination needs more transparency.**

Chair selection
   o This is likely to be the BOF Chairs, but not necessarily.
   o Chair selection is very important.  See separate message**.
 
Charter development
   o Chairs collaborate on charter development (consult mailing list).
   o More intense discussion of questions in BOF request/BOF review.
   o Does every milestone have someone with cycles associated with it?
   o Does every milestone match the purposes in the charter? Are all
     the milestones well-defined?  Does the charter conclude?

Charter review
   o IESG/IAB internal review.  This currently cc's the proposed chairs.
     Chairs often take up the discussion points on the mailing list.

   o IETF WG Review.  This currently involves ietf-announce and discussion
     on the IETF list.  We rev charters based on the discussions.  
       By tradition, IETF WG Review doesn't include the milestones.  
       Should this be changed**?
   
WG formation time lag**

This really varies a lot.  If the WG proposer does well with steps above,
there will be a BOF at one IETF, one cycle of reviews, and formation of the
WG by the next IETF.  (This may not be OK for IETF-wide overload, but I've
already overloaded this messages with topics, so enough :)

Allison