Re: [mpowr] SUMMARY: Give WG chairs authority to block documents?

["James Kempf" <[e-mail]>]

Margaret,

This is a good summary of part of the discussion. However, I think the
discussion has gone beyond what you've summarized, to include the entire
quality control process.

My own thinking on this has evolved in conversations with Pete, Dave, and
both Alexes. I now agree with the contention that WG chairs should not have
authority to block documents because spotting quality problems often
requires a bit of distance from the technical work. Thus chairs are unlikely
to have enough perspective to spot serious cross-area problems. The
situation with software QA is analgous. However, there does need to be some
kind relief from cross area review for the bulk of documents that the IESG
now does and which are fairly routine.

My understanding of where the discussion (which may have moved to the
solutions list) was at is that a review board like SIRS or ARTS would do
most of the reviews, and the IESG would only become involved if there was a
dispute or if the technology under development was new or resulted in a deep
change in existing technology. The WG would thus have primary responsibility
to resolve quality issues identified by the review board, with the chairs
co-ordinating, and most documents would be published without the whole IESG
becoming involved. In case of a dispute, the IESG would arbitrate between
the reviewers and the Working Group, and the document would be published
when the disupte is resolved to the satisfaction of the IESG (just as now).
In cases where the IESG chooses to do a review, the process would look
exactly as now, except the WG chair would be expected to co-ordinate more
actively during the phase in which quality issues identified in the IESG
review are resolved as was identified in your previous SUMMARY email. In
addition, reviews would be required earlier, to catch quality problems
before they propagate and require extensive rework, and ADs would be
expected to use reviews as a way of catching WGs that are drifting or not
making progress to help get them back on track, or terminate them if
attempts to help are not successful, more proactively than is currently the
case.

In summary, I think the discussion has uncovered how to reform the quality
control process in a way that would preserve the principles of community
based decision making and maintain the fairly strict vetting function that
we have today, but also offload the bulk of the review process from the IESG
to reduce their workload, rather than just the question of whether to give
WG chairs authority to block drafts.

Right now, I'm trying to write up my understanding of the group discussion
(also taking some points from the ART and SIRS drafts) in a draft which I
hope to have done shortly.

            jak

----- Original Message ----- 
From: "Margaret Wasserman" <[e-mail]>
To: <[e-mail]>
Sent: Friday, January 09, 2004 2:17 PM
Subject: [mpowr] SUMMARY: Give WG chairs authority to block documents?


>
> This message summarizes the third sub-discussion that we had
> under the heading of the MPOWR WG Proposal:
>
> (3) Should WG chairs have the authority to block documents
> due to low technical quality, insufficient review, etc.?
>
> This discussion took place under the MPOWR WG Proposal subject
> and also under the subject "Quality Control and that nasty
> A word".
>
> This discussion grew out of the discussion (mentioned in
> earlier summaries) about the distinction between shifting more
> responsibility to WG chairs and shifting more authority to
> WG chairs.
>
> In particular, Pete Resnick thought that he heard many people
> at the plenary say that giving WG chairs "the authority to say
> no" was a really bad idea.
>
> In response, Dave Crocker indicated that the focus on authority
> is misplaced.  What we really need to do is foster more
> collaboration.
>
> James Kempf pointed out that if we want chairs to be responsible
> for making sure that documents are well-reviewed and that the
> review comments are addressed, we will need to give them the
> authority to hold documents until these criteria are met.
>
> Pete Resnick stated his opinion that chairs don't need more
> authority to hold document in this case, and offered a definition
> of consensus that would allow WG chairs to hold these documents
> except in cases where the WG has specifically considered these
> questions and does believe that sufficient review has been
> obtained and the issues addressed.  James disagreed that
> chairs have this authority under the existing BCPs.
>
> A discussion ensued, the conclusion of which is that, although
> WG chairs have broad authority to define and determine cosensus,
> only ADs currently have the authority to hold a document
> because the document has technical problems or design flaws.
>
> Dave Crocker pointed out that this is correct and appropriate.
> The WG chair should never be able to overrule WG consensus.
>
> James Kempf commented that, in order for an early review system
> to be effective, the comments need to be binding.  In other
> words, WGs need to be required to address them.  James used
> the analogy of a QA department in a software company.
>
> Pete Resnick pointed out that the IETF is a open, consensus-
> driven organization and that this type of organization inherently
> gives more responsibility to its leaders than authority.  Some
> discussion ensued about whether the IETF actually is a hierarchy
> or not...
>
> Dave Crocker pointed out that in the IETF, the troops really
> are in charge.  Working groups have inherent power, whereas
> WG chairs and ADs only have derived power.  He later pointed
> out that explicitly assigning authority to one agent implies
> that the other agents do not have the responsibility.  In other
> words, if we may WG chairs responsible for document quality,
> that implies that the WGs are not responsible for quality.
>
> In response, Alex Rousskov started a thread with the subject
> "Troops vs.Superpower" which discussed the relative authority of
> the IESG vs. WGs.  This eventually led to Alex Conta and Robert
> Snively making a suggestion to move all technical authority out
> of the IESG.  Alex Rousskov, James Kempf and Pekka Savola
> disagreed, citing a lack of people willing to do completely
> non-technical IETF work.  Robert Snively suggested that this
> could be addressed by making the IESG a paid secretariat
> function.
>
> One useful take-away from this portion of the conversation
> is that it is important, as we discuss making changes to
> the responsibility/authority of different positions, to be
> mindful of what motivates people to participate in the
> IETF at each level.
>
> Alex Conta pointed that having WG chairs be involved in
> the technical work of a WG can present a conflict of
> interest.  Pekka Savola pointed out that this is often
> addressed by having more than one WG chair, so that no one
> has to serve in the chair role for his own document.  There
> was some further discussion about how/if WG chairs should
> be involved in the technical work of their WGs.
>
> Conclusions:
>
> I was unable to draw any firm conclusions from this portion
> of the discussion, except that it is clear that we do not
> have consensus among those involved in this discussion about
> either:
>
> (1) What level authority WG chairs currently have to block
> documents based on quality criteria.
>
> -or-
>
> (2) What level authority WG chairs should have to block
> documents based on quality criteria.
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
> _______________________________________________
> mpowr mailing list
> [e-mail]
> https://www1.ietf.org/mailman/listinfo/mpowr
>


_______________________________________________
mpowr mailing list
[e-mail]
https://www1.ietf.org/mailman/listinfo/mpowr