On 06/11/2013 11:02 AM, Ted Lemon wrote:
On Jun 11, 2013, at 1:52 PM, Doug Barton <dougb(_at_)dougbarton(_dot_)us>
wrote:
The flip side of that argument is that we don't want to assume
working groups are infallible, or more importantly not subject to
the groupthink phenomenon. Otherwise what is IETF LC for?
The IETF last call is for catching things the working group missed,
not for rehashing arguments that were beaten to death in the working
group.
As I understand it cross-disciplinary review is also an important
function of the IETF LC.
It is certainly possible for a discussion in the working
group to go one way, and then for the same discussion to come up in
IETF last call and go the other way, because the experts on the topic
were not included in the discussion, or because their advice was
inappropriately ignored.
Right ... for example the recent issue in regards to the potential
deprecation of the SPF DNS RRtype.
This does happen; unfortunately, when it
happens it often doesn't get caught in IETF last call anyway.
Then the process is faulty. :)
I should also point out that when I mentioned "groupthink" in my message
I was not doing so in a snarky and/or throwaway manner. The phenomenon
is very real, and IETF WGs are an ideal breeding ground for it on
several levels. It's _very_ easy for a group of humans to get into a
mutual confirmation bias feedback loop, and although we generally do
better on that point than a lot of groups would in similar
circumstances, having people from outside the WG review documents
provides (or should provide) a much-needed sanity check. In many ways
having people from outside the group provide a well thought out review
is _more_ valuable than the process of creating the document within the
WG itself (although obviously we need to give proper credit for that
creation).
Doug