ietf
[Top] [All Lists]

Improving IESG & IAB engagement when IETF comments are solicited

2014-02-07 10:22:58
For documents from working groups or individuals, it is typical -- and
generally considered to be required -- to have public review proceed in
a manner that fully responds to concerns that are posted.  This can be a
laborious and even frustrating process, but the community considers it
an important validation process for work carrying the IETF impimatur.

This diligent process is applied to working group Last Call and IETF
Last Call.  "Fully responding" does not mandate agreeing with the
concerns or ensuring that those with concerns agree with the outcome, of
course, but it does mean doing the work of responding in detail and
attempting to find resolution.


There is a decades-long pattern to the process of an IAB or IESG query
for community comments, about IAB or IESG actions that do not require
formal IETF rough consensus -- and sometimes even actions by them that do:

     1.  The IAB or IESG sends a note to the IETF mailing list,
         soliciting comments.

     2.  A variety of comments are made.

     3.  The IAB or IESG very selectively responds to comments, often
         only through one round of response.

     4.  The IAB or IESG publishes its decision

The details of the process can vary, of course.  Sometimes the process
looks very much like a Last Call from a working group, where the IAB or
IESG engages fully, with those making comments, so that the concerns are
fully explored.

But typically, the process is dramatically more constrained.

Simply put, comments often are entirely ignored, or receive at most a
basic response, with no followup.

The latest sequence about IPR issues and the Note Well demonstrate
exactly this problem.  The only exceptional aspect this time is that
quite a few people are pressing back for better engagement.

There is a basic difference between an obligation to discuss issues
diligently, versus an obligation to obtain some form of consensus.  The
former can be done without having the latter.  In effect, the former
calls for 'seeking' a consensus, while the latter demands achieving it.


Here's a different model the IAB and IESG should be following:

   1.  When a decision is not straightforward and reasonably warrants
       review by the IETF community, the IAB or IESG circulates its
       description of the issue, possibly noting the likely decision.
       (I believe that describes existing practice; no change is being
       suggested to this gating step.)

   2.  When comments are posted, each concern or suggestion that is
       raised should be treated as an item for explicit discussion,
       with a goal of trying to resolve it.  This is exactly the model
       typically followed for Last Call items and often followed within
       working groups, during regular development.  So we are all
       familiar with how to handle this process.

   3.  This imposes the burden of /fully/ tracking and discussing each
       concern /and/ of seeking resolution to the concerns that satisfy
       those raising them.  However, absent community consensus that
       the decision must seek IETF rough consensus, this process does
       not change who has final authority over the decision.


This sequence is substantially more work than what has typically been
done.


It's worth it.



d/
--
Dave Crocker
Brandenburg InternetWorking
bbiw.net

<Prev in Thread] Current Thread [Next in Thread>