Trusting the IESG to manage the reform process (was: Re: Doing the Right Things?)

[John C Klensin]

--On Tuesday, 03 June, 2003 06:41 +0300 
"john.loughney at nokia.com" <john.loughney at nokia.com> wrote:

> Actually, at the moment, I don't think it matters if there is
> a massive IESG conspiracy or not.  Trusting the IESG is
> irrelevant, IMO.  One feature that has really become apparent
> to me is that  the IESG is facing scaling problems.  If mails
> to ADs don't get answered within a reasonable time because ADs
> suffer from too much going on, I am not so sure adding to the
> IESG burden will be a good thing.

John, I've been arguing that the IESG is seriously overloaded 
and suffering from scaling problems and that we need to do 
something, or several things, to reduce the load for five years 
or so.  And I've been doing it fairly loudly.  So have others. 
Certainly, there are many signs and symptoms that look, from the 
outside, like overload -- slow responses to mail to ADs and 
apparently-unreasonable delays in document handling perhaps 
heading the list.  In that same period, we have gotten the 
tracking system: it may have taken too long to get in place, it 
certainly helps with transparency, but I have no idea what the 
additional reporting requirements actually do to workload and 
whether they are balanced (in load terms) efficiencies in 
keeping track of things.

But, the tracking system possibly excepted, the IESG has 
regularly, almost relentlessly, added to its load over that 
period.  In every single case, taken one at a time, there has 
been a good reason to add to load or to reject or ignore a 
proposal to reduce the load. E.g.,

	* If I work on, or partially rewrite, that document,
	rather than just bouncing it back to the WG or author, I
	can make it better.
	
	* This may never go to draft, so let's take the time to
	get those last 100 technical and editorial nits fixed.
	
	* Proposal to enforce benchmarks and give WGs more
	authority relative to ADs?  Really would not work and
	would eliminate vital cross-area review.
	
	* Proposal to reduce the number of WGs, or put a ceiling
	on them? Too ruthless, not really practical, takes
	important flexibility out of the system.
	
	* We aren't getting enough detail in some areas in
	documents submitted to us, so let's add another required
	section or more fixed requirements about how sections
	are written.
	
	* Need a liaison to XXX?  An IESG member should take it
	on, since only they really know what is going on in the
	work area.

	* Add a new AD, even temporarily, to deal with problem
	area YYY? Nope, the IESG is too big already, this would
	upset the balance of things, it takes too long to bring
	a new person up to speed, such a position is a bad idea
	anyway except maybe as a temporary position, and
	temporary positions are better filled internally.

There are more examples, some better and some worse.  The thing 
that is important about the ones I can identify is that every 
single one of them, examined by itself, is completely rational. 
All of the paraphrased assertions and conclusions in the list 
above are reasonable and valid, again, taken one at a time. 
Maybe SIRS will help and be the exception, but it would be easy 
to construct arguments against it too -- I can imagine the 
administrative/ policy burdens of actually maintaining/ 
administering the list and qualifications as being burdensome.

If one is amused by such things, it is possible to construct 
fanciful conspiracy theories to "explain".  I know enough of the 
IESG members to just not believe them.  And Occam's Razor 
suggests easier explanations.  I'm reminded a bit of "80 hours a 
week and loving it" bumper stickers.

My new-this-week working hypothesis is that, whether they appear 
objectively overloaded or not, we need to stop making decisions 
based on the assumption that they are.  We can't reduce the 
workload against their desires to take on extra tasks, their 
inability to refuse new ones, or their inclination to expand the 
IESG's role on a "this is important, someone needs to do it, and 
we are the obvious answer" basis.

If an AD isn't being responsive, tell it to the Nomcom, possibly 
with advice that they pick someone who is strongly committed to 
turning away more work, rather than willing to add it.  If you 
can't wait that long, and _really_ think it is that bad, think 
about recalls. But, IMO, our saying "we can't think about doing 
it that way because the IESG is overloaded and/or suffering 
scaling problems" is pointless until and unless the IESG is 
demonstrably ready to start making hard decisions to reduce 
those problems... And we had been be ready to support them, 
strongly, when they do start making those decisions, because, 
for any given one of them, _someone_ is going to be really 
unhappy.

    john