--On Thursday, June 27, 2013 11:07 -0400 Michael StJohns
<mstjohns(_at_)comcast(_dot_)net> wrote:
...
But that's still problematic. The current rules basically
give any company who provides >= 30% of the Nomcom volunteer
pool an ~85.1% chance of having 2 members (sum of all
percentages from 2-10 members), a 12.1% chance of having 1 and
a 2.8% chance of having 0.
And, of course, a few related companies or one with subsidiaries
that it claims are independent and a desire to game the system
can easily do so today to produce a highly-likely larger number
of members, perhaps even a majority. We are clearly vunerable
in that area but, perhaps inevitably, have gone down multiple
ratholes as soon as we try to tighten the specifically
anti-capture rules further. In particular, trying to
second-guess a company's assertions about what entities are
independent of it is just fraught with problems even thought
such assertions can be used to subvert any, or almost any, "no
more than X from one company/interest" rule.
I believe the proposal as stated would further exacerbate that
problem - not for a given company, but for pretty much locking
small companies and individuals out of the Nomcom. Once
scenario for this - both benign intentions and non-benign - is
that a company instead of sending one person to all the
meetings starts rotating the opportunity to attend the IETF
among a number of people - say 5. So instead of the potential
of say 30 volunteers from one company, we now suddenly have
150. And me with my single person consultancy - still only
has 1 slot to volunteer.
Yes, but...
(i) Some reasonable set of "demonstrate clue" provisions, if
adopted along with the "demonstrate participation" ones, would
tend to counter that, and might even reduce the available 30 (or
whatever number one picks).
(ii) A notion of "some remote participation is ok if you can
demonstrate clue/ groking the IETF in some other way" actually
helps to protect your ability to volunteer if you suddenly
missed a few consecutive meetings and that of other one-person
consultancies (including, in the interest of full disclosure,
me) and academics like Brian) who have discovered that, as
in-person attendance rises, they need to be selective about face
to face attendance. If one (and, equivalently, one's one-person
consultancy) can't volunteer, the likelihood of being selected
are zero (and really easy to calculate).
(iii) Unless all of those 150 people were actually participating
actively and contributing to IETF work, the company's actual
ability to influence the standards process would go down.
While it would be good to have more people involved, it would
be bad in the ways in which larger companies could game the
system.
While I agree, I think it would be really unfortunate if we
discarded opportunities to get a broader spectrum of people
involved because of fears that come company or companies might
misbehave. I think at least two things work in our favor in the
latter regard. We can try to educate participants and their
organizations about how little marginal commercial advantage
they would gain by, e.g., getting an extra person on the IETF
and be sure that our other procedures and safeguards reinforce
that. For example, the reputational harm a company, to say
nothing of the individuals involved, would suffer from a few
recalls based on collusion in support of company positions would
far exceed any possible advantages from having more people in
leadership positions. I think there is also evidence of courts
and competitiveness/ antitrust authorities being _very_
unsympathetic to organizations who try to subvert open standards
processes. They, of course, also have the resources and
authority to untangle relationships that we do not.
best,
john