--On Monday, February 19, 2018 09:38 -0800 joel jaeggli
<joelja@xxxxxxxxx> wrote:
>...
>> Joel,
>>
>> At the time 2418 was written, areas with more than one AD
>> existed, but were not the norm, and, IIR, we had never had one
>> with three ADs. Also, while WGs that had existed for years
>> existed (a few of those continue to exist near 20 years
>> later), they are now much more common, making multiple
>> responsible ADs over the lifespan of a WG almost inevitable.
>> If there is one one AD in an area, the Nomcoom at least has
>> the opportunity to consider effects on continuing WGs when
>> deciding how to fill an open slot. In multiple-AD areas,
>> when ADs turn over, WGs are typically reshuffled among the
>> new and returning ADs. At least in many cases, those
>> decisions occur without community visibility and without any
>> explicit discussion with the affected WGs or even their
>> chairs.
>
> The IESG believes it has the power to divy up work among the
> available members. We have plenty of existence proofs dating
> back decades.
And there is no disagreement, at least AFAIK, about that
"power". I do believe that, especially in the light of both the
industry consolidate you mention below and other factors that
are narrowing support for IETF-style standardization work, it
may be desirable to be a little more open about the divvying-up
process. As I tried to say earlier, I am not claiming that any
abuses have occurred. The suggestion is only about avoiding the
appearance of abuse in the future. In any event, if abuses had
occurred, they would be a separate question,
>> Also at that time, we tended to get a little anxious when the
>> same company (or its near-affiliates) was employing or
>> supporting multiple members of the IESG and/or IAB. By my
>> rough count, today we have three organizations with two IESG
>> or IAB members each and one with four. There was also a
>> strong cultural assumption in the IESG that having an AD
>> responsible for a WG whose chair was employed or supported by
>> the same company was a seriously bad idea. I gather that
>> assumption has been relaxed as well, especially with WGs with
>> co-chairs.
>
> Getting anxious is the nominal state, that means we're on the
> lookout for the appearance of conflict of interest.
Yes, see above. I am only suggesting that, if things continue
to evolve as they appear to be evolving, we may need to be even
more careful and, in the process, look for ways in which
procedures and transparency may need to be reviewed to be sure
we are still on track, noting that "I recuse myself and will not
vote on this draft but may still participate in general
discussions about the associated issues and context" or "I have
written and am pushing this particular document but someone else
will sponsor or shepherd it in the IESG so everything is fine"
may not be sufficient long-term and that, were the number of
people on the IESG with affiliations to the same organization
("affiliation" may including relationships other than direct
employment) were to rise significantly, it could complicate
document reviews and appeals.
> The Industry that contributes people and time to the IETF is
> consolidating. Eventually it will be gone and we'll have to do
> something else; Until then it's harder to throw willing and
> able contributors under the bus because of the incoveience of
> their current employer then in the past.
Absolutely. However, I'm raising this issue at this time to
minimize the risk of the IETF itself going under that bus and
perhaps to suggest that the time to start thinking about
"eventually" and how me might need to consider and adopt
relevant "something else" procedures is probably well before
some unpleasant event occurs rather than in reaction to it.
That doesn't mean we need to drop everything and panic, but it
does suggest that, if your analysis (with which I agree) is
correct, it may be time to start thinking more about the issues
and risks rather than claiming that, because there have been no
serious problems yet, everything is fine.
>...
>> At least from my point of view, part of the problem is that
>> while the IETF has been paying very careful attention to
>> intellectual property issues, there has been too little
>> concern about that appearance of conflicts of interest,
>> concentration of decision-making in particular industry
>> actors, and antitrust and competitiveness issues.
>
> They're not solvable problems.
There we disagree. Even if they are not solvable, it seems to
me that is not adequate justification for ignoring them as
uninteresting. I note that many other standards bodies have
procedures in place to deal with issues of this sort, have
adopted specific guidance for handling situations that might
appear questionable as well as guidance for what decision-makers
with perceived conflicts are expected to do to avoid improper
influences, etc. As with IPR issues, there may be cases in
which having clear guidelines may be more important than the
specifics of those guidelines. And perhaps we should be
thinking about when and how to to start drawing an industry
process together to discuss the issues and start developing a
plan -- similar to what we did years ago for the first important
discussions of what became our IPR policies.
>> Or answer to those concerns has often
>> been that everyone participates as an individual so there is
>> no problem but, especially in situations in which "follow the
>> money" is an important maxim, that claim would not survive a
>> laugh test and "I recused myself by not voting on that matter"
>> (even though I might still participate in informal discussions
>> and give advice about it) wouldn't last much longer.
>
> If we produce bad work product because we've been railroaded
> down an particular architectural path by a bad actor, how is
> that distinguishable from previous outcomes?
IANAL, and I don't think you are either, but I imagine that, if
the IETF were to find itself in the middle of litigation (or
worse) charging that the decisions we made were driven (or even
just distorted) by a conspiracy to distort the marketplace to
favor a particular company or cluster of companies, we would
have no trouble at all distinguishing that from prior outcomes
or technical decisions.
> The ability to construct a better submarine is not in much
> doubt.
>
>> These may be issues that the IETF will continue to ignore.
>> If we extrapolate from the past, ignoring them has been a good
>> strategy because there have been no serious ill effects of
>> doing so. However, neither the WG-AD relationship nor the
>> suggestion that Nomcoms should be able to pay attention to
>> how particular AD candidates have handled or might handle
>> potential conflicts of interest strikes me as anything near a
>> non-sequitor.
> If your candidate pool consists of one or two individuals for
> a seat the granularity with with respect to choice is limited.
> Nomcom interviews / questionnaires I've been subjected to have
> probed how ADs address conflicts, especially in areas where
> they are the principle contributor. There is also a strong
> bias favoring incumbents unless they've proved somehow
> unsuitable; given that; about half the time they question comes
> down to how well did they do it, rather than how might they.
I don't want to try to dissect the selection process here -- all
I was suggesting in that regard was that it was important that
the Nomcom have adequate information available and that they be
both able and willing to evaluate it. But I would suggest both
that there have been occasional situations in the past in which
a Nomcom has apparently ignored bad behavior, had insufficient
information, or considered the behavior and concluded that, on
balance, a particular candidate is nonetheless the best choice.
We have also had discussions in the past about whether a number
of terms that reflects the incumbent bias you suggest above and
that are limited only by exhaustion of an incumbent or
withdrawal of support is necessarily a good idea, and so on --
discussions that have led nowhere (of course, that might have
been the right outcome).
best,
john
