Harald,
You asked, and I responded, privately about this, but, since you
have identified it as a consensus request, I suppose I should
restate my concerns in public. Unfortunately, some of this
revisits the area in which I think Sam (and others) and I have
agreed to disagree -- I don't know how you are going to resolve
that.
Inline below...
--On Friday, 28 January, 2005 15:02 +0100 Harald Tveit
Alvestrand <harald(_at_)alvestrand(_dot_)no> wrote:
In case some people did not see this in the flurry of messages
on the thread, here is Margaret's proposed text with Leslie's
edits, incorporated as best I can.
I also tuned grammar slightly in some places.
3.5 Review and Appeal of IAD and IAOC Decision
The IAOC is directly accountable to the IETF community for
the
performance of the IASA. In order to achieve this, the
IAOC and IAD
will ensure that guidelines are developed for regular
operational
decision making. Where appropriate, these guidelines
should be
developed with public input. In all cases, they must be
made public.
Additionally, the IASA should ensure that there are
reported objective
performance metrics for all IETF administrative support
activities.
Back when I was actively doing political science, the belief
that everything could be reduced to objective and quantifiable
terms (the latter is what "metrics" means; if it isn't what was
intended, some other word should be chosen) statements like this
were described as "physics envy". The statement would be
reasonable if "whenever feasible" or the equivalent appeared
there somewhere -- we _can_ evaluate the IAOC on its
interpretation of "feasible" and how far they are willing to go
to satisfy the needs or curiosity of the community.
...
on the nature of the review request. Based on the results
of the review,
the IAOC may choose to overturn their own decision and/or
to change their
operational guidelines to prevent further
misunderstandings.
This doesn't give the IAOC the option of saying "no, you are
wrong [because...], and we aren't going to change anything".
Combined with other text above, that would imply that any member
of the community can force the IAOC into either changing a
decision or changing the operational guidelines. The IAOC must
be able to say "no you are wrong". If must even be able to say
"you have raised fifteen objections in the last 30 days, all of
which have been turned down by us and everyone in the appeals
chain, please go improve you sand-pounding skills".
If a member of the community is not satisfied with the
IAOC's response to
his or her review request, he or she may escalate the
issue by appealing
the decision or action to the IESG, using the appeals
procedures outlined
in RFC 2026 [RFC2026]. If he or she is not satisfied with
the IESG
response, he or she can escalate the issue to the IAB and
on the ISOC
Board of Trustees, as described in RFC 2026.
I apologize in advance for digging up this dead horse for
kicking purposes, but I do not believe that an _appeals_
procedure that runs through the IESG is appropriate. First, it
increases the IESG workload and, while our reasoning and the
details differ, I agree with Avri that shifting work back to the
IESG is not a good idea. Second, the IESG is not a collection
of general-purpose representatives of the community. It
membership is selected to manage the _standards process_ and, as
a member of
the community, I want their selection to continue to be focused
on selecting people whose skills are optimal for that role. I'm
not happy about the IAB being thrust into the appeals role, but
I can more easily see that in terms of workload and make a
"general purpose representative" argument for them. The fact
that the IAB and IETF Chairs (who are, by definition, implicated
in any consensus IAOC decision-making) are the leaders of those
groups (whether formally recused or not) makes an appeal model
even more complicated. To the extent that any appeal addresses
substance, rather than compliance with procedures (and anything
that might be overturned necessarily does so), all of my old
comments about training in finance and executive management
apply here as well.
The reviewing body (IESG, IAB or ISOC BoT) will review the
decision of the
IAD or IAOC
to determine whether it was made in accordance with
existing BCPs and
operational guidelines. As a result of this review, the
reviewing body
may decide to initiate the required consensus process to
change the BCPs
governing IAOC actions.
This is not quite right. IETF "public consensus processes" are
initiate under the rules of 2026 and associated documents. We
don't initiate them until we have a document on the
table. _Anyone_ can put a document on the table, with or
without a prior IESG/ IAB/ ISOC BoT decision. The decision,
presumably by the IESG, as to whether to Last Call such a
document, organize a WG to discuss changes or alternate
versions, etc., presumably gets made or not made on the basis of
the merits of the proposed revision and a sense of whether the
community wants change, so the outcome of a prior appeal doesn't
count for much, even though a record of appeals being filed
might. So, if that is all those bodies can do, then the appeal
text is pointless.
They may also advise the IAOC to modify existing
operational guidelines
to avoid similar issues in the future and/or may advise
the IAOC to
re-consider their decision or action.
This is actually weaker than the previous text, in which
_either_ the IAB or the IESG could _require_ that the IAOC
review and, if appropriate, reconsider the decision or action.
I can live with it, but I'm astonished if Sam, Avri, etc., can.
In exceptional cases, when no other recourse seems
reasonable, the IESG,
IAB or ISOC BoT may overturn or reverse a non-binding
decision or action
of the IAOC. This should be done after careful
consideration and
consultation with the IAOC regarding the ramifications of
this action. In
no circumstances may the IESG or IAB overturn a decision
of the IAOC
that involves a binding contract or overturn a
personnel-related action (such
as hiring, firing, promotion, demotion, performance
reviews, salary adjustments,
etc.).
My preference, per your other comments and the paragraph below,
is that we lose the above paragraph. But, if we leave it in, I
suggest that we need to define "exceptional cases" or the
mechanism by which it is determined. The issue here is very
much the same as the argument raised by Pete and others that the
ISOC Board should not be able to back out of this
agreement/model by a simple majority vote at a single meeting.
The determination that a case is [sufficiently] exceptional
should take, e.g., extraordinary consensus, announcement to the
community and an opportunity for comment, and, more important,
identification of the individuals who favor and oppose declaring
that exceptional circumstances exist so that they can be held
accountable.
In another message, I have suggested removing the last
paragraph - but it's not a show-stopper for me to leave it in.
Comments?
I'm not sure we are getting closer.
john
_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
https://www1.ietf.org/mailman/listinfo/ietf