[Icar] Re: [Solutions] Summary of Discussion on Reforming IETF Quality Control Process

Hi Margaret,

That was quick! Here are a couple of quick answers, but if you think this
draft might have legs, I'll get an issues page up on it.

> What are the criteria for deciding whether a particular document
> requires full IESG review?

It is up to the IESG to decide that. The WG can ask for a full review if
they think it might be necessary, but the IESG makes the decision. Also, the
shepherding AD can ask for a full review if s(he) thinks its necessary. In
general, I think the IESG would want to review drafts from WGs that are
proposing new technology or major changes in old technology, but I did not
think it would be such a good idea to be that explicit because it would
decrease the attractiveness of serving on the review board, and, as a
practical matter, the IESG may want at some point to have the review board
do a review on some draft that proposes new technology if it's a small
change or a major change in old technology if it isn't all the different or
if the WG is composed of people who are known to be reliable and world class
experts in the field.

>And, when would this be decided -- at
> charter time, when the document is accepted as a WG item, during
> WG Last Call, when the shepherding AD reviews the status of the
> document, or at some other time?
>

In the normal case, as soon as the draft showed up as a work item on the
WG's goals list, which is what the process draft should say because some WG
drafts are added after the charter has been written. A review would have to
be done before the draft became a WG draft in any case. The idea here is to
inject a note of predictability into the review process, so the chair, WG,
and shepherding AD know exactly what to do when the time comes for the
review.

However, in exceptional cases, typically when the shepherding AD has
detected that the WG is off track or has reason to believe that the
solutions being discussed by the WG may be impractical or otherwise flawed,
the shepherding AD can call for a review and include an IESG review of
documents, if talking with the WG fails to reorient the WG.

> In dispute resolution, you seem to indicate that the shepherding
> AD will take all disputes to the full IESG.  Currently, the
> shepherding AD (or two ADs in an area) make an attempt to resolve
> any dispute before it is brought to the full IESG.  Is it your
> intention to change that?
>

Yes, I believe that is probably a better idea and I will add that to the
document. No reason for full IESG intervention if the WG and reviewers can
come to an accommodation with the AD's help.

> Do you have any sort of transition plan for how we could move
> to this approach?  Are there useful experiments that we could
> run to see if it would work?
>

One way to start would be to pick a few WGs and give it a try. The list of
SIRS are obvious candidates for initial Reviewers. Obviously, if the process
is applied to some WGs that exist initially, they will not be able to set up
the review process when they are initially formed. I'd say picking groups
such as DHC where many of the drafts involve proposals for new options or
other small changes would be one way to test the review board process, I'm
not sure which ones would be appropriate for the new/deeply changed
technology case. The experiment could be run for, say, a year and closely
monitored to see how much work it really did eliminate for the IESG and
whether it really did free up more time for ADs to spend working with their
WGs, then tuned as appropriate or even discarded if it isn't working.

            jak

_______________________________________________
Icar mailing list
Icar@ietf.org
https://www1.ietf.org/mailman/listinfo/icar