Re: "Discuss" criteria

[Lisa Dusseault <lisa@xxxxxxxxxxxxxxxxx>]

On Jan 1, 2007, at 4:49 PM, John C Klensin wrote:

... regardless of what we do, some

of these will ultimately hit the IESG during final review (as

well as agreeing that the Kerberos WG is a poor example).

However, I think your comment could be construed as a little too

accepting of the situation.  From my point of view, any time

that one of these situations is not discovered before Last Call

is an indication of a failure in the system, a failure that the

relevant ADs should be examining carefully in the hope of

preventing future failures of the same or similar nature.

    john

Indeed, although surprise issues are not completely avoidable, we should look on each one as an opportunity to ask "how could we have caught this one earlier" (just like bugs in software found after checkin). 

So... how do we do better generally?  One possible approach:  I'm hoping that creating outside-expert review process for early review will help with this a bit.  It might not help much but it's not a very expensive experiment to run.  We've already got a few groups of experts (directorates, area review teams) who've volunteered to do early review.  Last week I posted to the WG chairs list asking who wanted to volunteer their documents to go through this a little more formally, with a bit of tracking of what reviews are requested and received.

Lisa

_______________________________________________

Ietf@xxxxxxxx
https://www1.ietf.org/mailman/listinfo/ietf