--On Friday, 10 October, 2014 11:37 -0400 Michael StJohns
<mstjohns@xxxxxxxxxxx> wrote:
> At 03:10 AM 10/10/2014, John C Klensin wrote:
>...
>> I think we have a little bit of a philosophical disagreement
>> that has a lot to do with what the IETF is about and much less
>> to do with this particular case.
>
> I'm going to pseudo-top post here. I don't think of this as
> so much a disagreement in the philosophy of how flexible we
> should be, but more a disagreement about when it's expedient
> and appropriate to throw the rules away. Most of the stuff
It seems to me that we aren't going to convince each other or
much of anything else, so this will be my last posting on the
subthread of whether the IESG can do this. I also believe the
Nomcom can make one-year appointments. It has certainly done
so: Michael's note and see Barry's note that lists recent
precedents.
FWIW, I also believe that the Nomcom can leave a position that
the IESG or IAB requested open. If nothing else, a Nomcom can
say, "unable to fill that on schedule, so sorry" -- as was done
with Transport not that long ago-- and then never quite get
along to filling it. That could get quite ugly if the community
did not understand and accept the reasons but we have no
procedure that I'm aware of that would allow forcing the Nomcom
to do anything, no procedure to force the Nomcom to explain any
of its personnel decisions to the community, much less get
community consensus for them, etc. Under most circumstances, we
see that degree of independence as a necessary safeguard. If
one believes strongly that the Nomcom MUST do something like
fill all requested positions, it is a little awkward.
> we've come up with has a deliberative component to it -
> "figure out what you need, socialize it and get consensus"
> coupled with procedural items ("normal AD appointees take
> office during the First IETF"). If you follow the rules,
> you mostly get a predictable outcome and predictable times.
> If you want to bypass the rules, then you *still* need to
> "figure out what you need, socialize it and get consensus",
> but not just for the change, but for the change in procedures.
Per the above, I don't even see the alternatives that have been
proposed as changes in procedures, much less ones that are so
basic as to constitute "throwing the rules away".
While the tendency has been for the IESG to expand, adding areas
and moving, one area at a time, from one AD to two, we've also
eliminated them, usually but not always either coincident with a
Nomcom cycle or by shifting the AD into another area. Remember
"OSI Transition"? "Applications Support"? (My memory may not be
good enough to remember exact names, but the Areas definitely
existed.) IIR, we even had Operations and Management separated
at some point.
> In the instant case we got "The IESG has decided to close out
> the APPs area - discuss. Oh and by the way, while the
> discussion is taking place you guys over at the Nomcom should
> ignore the open slot in APPs in your deliberations." The
> first piece of it is fine; the second less so. As has been
> evident by the discussion, the form of the IESG in fall of
> 2015 is still up in the air; leaving me to wonder if there's
> something else going on that's not been made public.
Again, at least as I read the procedures, the IESG has the
right/authority to tell the Nomcom what slots to fill and,
procedurally, this seems consistent with that authority if
uncomfortably late. I also don't see a lot of difference in
practice between appointing an AD to a one-year term and
appointing an AD for two years with the explicit understanding
that the Area is probably being reorganized and that, while the
newly-appointed AD will get a voice in that discussion, he or
she may be expected to step down or shift into work with another
designation before the term is up. See below.
It is actually not even clear to me that the IESG is required to
see community consensus to reorganize itself. I'm quite sure it
has reorganized areas in the past by announcement, without any
effort to get such consensus or even a discussion.
None of the above suggests that I think either this move or the
timing are wise or in good taste. Those are separate issues. I
just don't see a huge procedural problem, much less one that
rises to "throwing away the rules".
> I've no idea how far the Nomcom is along the process. AIRC,
> Nominations were supposed to be in as of 8 October or right
> about when this got started.
Another example, in a way. The only way that the close of
nominations interacts with this is that, if there are applicants
for Apps in the pot and we continue on the course selected by
the IESG. then the Nomcom says "in combination with the IESG"
(or "at the direction of the IESG") "we aren't filling the Apps
slot. Thanks for your interest in the position". In part
because I believe that the Nomcom has the ability to go
recruiting and even arm-twisting of people who didn't volunteer
pre-the nominal nominations cutoff, I believe they could also
interpret a nomination for Apps as a potential nomination for
some other Area, go back to the individual, and say "you were
nominated for the Apps slot and there isn't one, would you be
willing to serve in XYA". You might disagree about that too,
but, again, there is precedent for candidates being appointed to
particular positions who where not nominated prior to the
nominal cutoff so, if there were a procedural violation, it
would be old news.
>...
>> The IETF, by contrast, tended to focus on identifying the
>> right things to do and getting them done.
>
> Yes but. There's a difference on getting documents through
> the process (and it's unclear we're doing better there than we
> used to be) and all the rest of the cruft that makes up an
> organizations operating system. I'm not saying that applying
> "do what's right" to the operation of the IETF is necessarily
> wrong, but that's the place were we touch the legal systems of
> various jurisdictions. Not following our own rules is one way
> to get in trouble.
I completely agree and worry about that a lot. I am, for
example, very concerned when we make standards-track decisions
based on individual submissions where there is not clear
evidence of community review and consensus (not just silence by
everyone but the author and a small circle of supporters with
personal or corporate connections). But this particular issue
seems less significant, less of a deviation, to me than it
apparently does to you. I also believe that almost every aspect
of it, even if not the combination, has precedents so it
wouldn't put us in any stew we aren't in already. Of course,
if there were any evidence that either the decision or the
timing were motivated by personal or corporate considerations
that have not been exposed to the community, that would be a
rather different matter.
> All of the above said, my large objection to this is that it
> seems rushed, and not well thought out with possible hidden
> implications and unintended consequences.
Again, we are in complete agreement on that subject.
Personally, I'm particularly concerned about positive feedback
influencing work choices, e.g., the chilling effect on someone
thinking about bringing new work into toe Apps area or even
keeping work within the IETF that is struggling. Area
designations and specialties do make a difference. Anyone who
needs an example should consider the course of DNS work
(including IDN work) since the DNS was shifted from being
considered as an application, into an applications
infrastructure area, and then into Internet. It is not clear
that the work taken up and many of the decisions made would be
different, but some of the outcomes certainly would have been.
Because it seems directly relevant...
--On Friday, 10 October, 2014 10:33 -0500 Pete Resnick
<presnick@xxxxxxxxxxxxxxxx> wrote:
>...
>> We really do not need to argue
>> about whether one-year appointments are possible because, if
>> the IESG decides to eliminate an area or AD slot and the
>> relevant ADs are so attached to the jobs that one or more is
>> willing to resign for the good of the community and to free
>> herself from a non-existent position, I suggest that we have
>> far greater problems than the number of years for which a
>> Nomcom can make an appointment.
>
> Assuming you meant "not willing",
yes, sorry.
> I think you're being a bit
> unrealistic, or at least not being nuanced enough, about this.
> For better or worse, the way we select folks for the IESG now,
> we ask for a serious two year commitment, one that normally
> involves talking to employers and re-arranging life for quite
> some period of time. Even if *theoretically* a given AD would
> be willing to resign for the good of the community, there are
> going to be cultural and social pressures on the IESG not to
> "re-organize one of their own out of a job". I'd like to think
> we weren't susceptible to such influences, but I think it's
> there in the background. Furthermore, it's not really fair to
> ask someone to make a commitment in such a nebulous state.
> Taking this opportunity (when you have an incumbent AD
> stepping down anyway from an area that currently has a
> light(er) work load and is clearly going to be part of the
> re-organization) to not fill the position until the re-org is
> done makes the most sense to me.
First of all, to the degree the situation is as you describe
(and it probably is) the subthread that Stewart started is
extremely relevant. The model we are using today is more than
20 years old. A lot has changed in that time. Unless we want
the IESG role to explicitly become a career choice [1] and a
commitment decision that can be made only by large organizations
and the independently wealthy, it is probably time to reconsider
how we organized ourselves and try to get work done.
That said,
> Trying to re-org and drop an AD later if we decide we need to
> is not a good plan forward. In my judgment, it simply won't
> work.
Interesting, because there is lots of precedent for it. I don't
remember anyone stepping down mid-term, but I remember creation
of special areas that expire with the individual, assignment of
someone as an extra AD to another area, and similar moves. My
bigger concern about putting an AD in for a short term is the
issue that Ted mentioned: the start up time is pretty painful.
On the other hand, as was pointed out in an off list comment,
perhaps there is a former AD or someone else with a head start
on the learning process who would be willing and able to take on
a short term commitment... possibly even one who couldn't commit
to the long periods and workload you cite [2]. One doesn't know
if one forecloses the possibility of asking.
In this particular case, given an extrapolation from the
combination of your note to the Apps Area about stepping down,
your obvious interest in how the IETF is organized and reaches
consensus, and dropping Apps workload, it might be sensible to
create a temporary "fishy" area (i.e., an offspring of a WG I
once chaired), put you in charge of it on a ramp-up transition
basis, and then persuade you to put in an extra six months to
shepherd and complete that work. That would make the timing
about right, would eliminate some of my concerns about WGs
suffering from even further reduced attention, and so on.
I don't know if it is a good idea or not. But it is yet more
evidence that, if the IESG actually considered the full range of
options, they haven't shared those considerations with the
community.
john
[1] A choice that, in many cases, makes sense only for
professional standardizers and not for architects, designers,
developers or those we traditionally think of as your primary
participants.
[2] For the record, I would absolutely not be a candidate. Too
old, insufficient support, feeling pretty burned out about the
IETF, and generally the wrong person for the job. So I'm not
shopping for a position, just identifying a possibility.