ietf
[Top] [All Lists]

Re: Appointment of a Transport Area Director

2013-03-03 08:01:14
--On Sunday, March 03, 2013 12:50 +0000 "Eggert, Lars"
<lars(_at_)netapp(_dot_)com> wrote:

The likely possibility is that many qualified people failed to
get sufficient employer support to be able to volunteer. It's
at least a 50% time commitment.

Yes.  And with emphasis on "at least".   See below.


--On Sunday, March 03, 2013 07:56 -0500 Eric Burger
<eburger(_at_)standardstrack(_dot_)com> wrote:

The 50% time commitment is an IESG-imposed requirement. If
that is really the problem, we have had areas with more than
two ADs.

Eric,

You say that as if the IESG sat down and said "let's make up a
50% requirement" and could have as easily said "let's make up a
25% requirement". That number is supposed to be IESG's best
estimate about the time commitment required, not a requirement
they have somehow set.   While I've often been critical of the
IESG, I have no reason to believe that they fudged the number
upward.   From observations of how much time and energy several
ADs put in, I would, if anything, suspect that a 50% estimate is
too low.

You also know better than to believe that adding another AD slot
would have much influence on this.  First, ADs spend significant
time on areas other than their own.  Unless the review rules
were changed so that only one AD from a given area even needed
to look at a document or perhaps even participate in IESG calls
and discussions, increasing the number of ADs in an area doesn't
help with that source of time commitment.   Possibly dividing up
the area's working groups three ways rather than two would help,
but I note that Transport doesn't have a disproportionately
large number of WGs.

Whether you think the Parkinson or the Brooks analysis are the
most relevant, work expands to fill the available time.

I do think there is a problem with that 50% number (or the large
number that I suspect is more realistic).  I've explained that
concern several times on this list (and more times in various
process WGs).  In essence, I think that IESG members who take
their responsibilities very seriously (a good thing, up to a
point) and who are able to spend most of their time on the IETF,
have, together, gradually taken on more tasks and
responsibilities for the IESG. Those tasks, in turn, have made
the job bigger.  I believe that the IESG should be managed, and
manage itself, so that each new task and each new BOF or WG, is
evaluated on the assumption that IESG time is a scarce resource
and that increased workload is not acceptable.  And I believe
that sunset provisions should be applied to existing tasks and
procedures with an overall goal of getting the IESG job back
down to the level that it could be done alongside a realistic
day job with real design, product, or implementation
commitments, not instead of one (once upon a time, that was the
norm).   

There are other ways to accomplish the same thing and to do so
more quickly.  For example, while I'm not recommending it, one
could, in principle, separate WG and IETF management and
"steering" functions and document review and approval functions
off into separate bodies.  But no one seems to even want to
discuss that one, perhaps for good reason (or not).

I'm certainly not either the first or the only person who has
suggested guidelines or mechanisms that would make the above
feasible.  It is perhaps interesting that none of the various
proposals have even been allowed to make it into IETF Last Call.


I believe strongly that a smaller size of commitment would get
us more applicants for _all_ AD slots, if only because it would
make the commitment from an employer or sponsor significantly
smaller.   But accomplishing that requires that the IESG reverse
the trend toward ever more responsibilities and an ever larger
job and time commitment.  Nomcoms could help with that by
selecting IESG members who are committed to a reduction in IESG
functions and workload as a goal (something I said every year
until I concluded that no one was listening).  But there is no
quick fix like juggling numbers or adding ADs: the first of
those would accomplish nothing and risk mid-term resignations
and the second would be likely to, in the long run, make things
worse.

The only thing that surprises me is that this is coming up in
only one particular area.  Given the implications of going to an
employer and saying what amounts to 

        "How about you help the IETF out by giving up the
        services of someone whom we presume is a highly-skilled
        designer for, realistically, four or six years, while
        continuing to pay him (or, better, her) full salary and
        supplying essentially unconstrained travel and other
        support resources.  During that time, you can certainly
        use her as an internal consultant or equivalent, but
        don't expect any real company work to get done.  And, by
        the way, you also have to protect that person's job,
        role, and advancement path within the company or no one
        sane would take the position."

...we should probably be pleasantly surprised that anyone, or at
least anyone who is professionally competent and useful, is ever
allowed to volunteer for any AD job these days.

      john