Re: 1601bis: The Charter of IAB

Date view Thread view Subject view Author view

From: John C Klensin (
Date: Sat Feb 12 2000 - 08:29:13 PST

--On Saturday, 12 February, 2000 14:40 +0800 "Rahmat M.
Samik-Ibrahim" <> wrote:

>I also believe what Klensin wrote is still valid:
>[... 22 Feb 1996 ... skipping voting out of existence, fine
>lunches and dinners, and irresolvable controversy
> pronouncements...]

My, my, do I love being quoted out of context on a comment that was
made four years ago, and was intended as humorous rather than a
serious comment on the process. I don't remember the comment or its
context, but strongly suspect from the partial sentence excerpt that
it was _never_ "valid" except as a sarcastic and/or cynical remark,
made in passing and intended to mean the opposite.

Rahmat, this is one of the difficulties with the path you seem to
want us to go down. Except when formally hearing an appeal and,
possibly in a very small number of other cases (I can't identify any
at the moment, but don't want to get into a nit-picking argument with
you about whether they exist), the IAB isn't a formal deliberative
body. And, even when we face those more formal situations, we are
not, as we pointed out recently in another context, a court of law
and will not try to behave like one. Instead, we operate very
informally, as a collection of people trying to collectively
understand issues and help the IETF move forward. All of those
people bring strong technical skills and experience to the table,
rather than being professional or amateur lawyers or the like. If
you don't like the composition of the IAB in terms of the individual
profiles and personalities, that is a Nomcom issue, not a charter or
procedural one.

In working informally --with each other and with the IETF participant
community (especially those who are making technical contributions)--
we also tend to be human. Under stress, some of us even become
cynical or engage, informally, in what we call in the US "black
humor". So, as you must know, do several other professional
population groups: this shouldn't be any surprise or cause for
concern. But that informal process is precisely what permits us to
work as a team, in spite of different backgrounds and perspectives,
and quite often come up with results that are clearer and better than
any of us could have produced alone. Tying us up in procedures, or
requirements that we expose and document every word spoken or
written, would, at a minimum, cripple our ability to get anything
done. That would, of cousre, prevent our doing harm by preventing
our doing anything, but I, at least, don't see that as in the best
interests of the IETF or the Internet generally.

I think those of us who have been following your notes over the last
few years understand your concerns but generally do not agree. As, I
believe (haven't checked), the only member of the current or upcoming
IAB with significant professional training in the social and
behavioral sciences, I am probably more sympathetic than most to your
desire for a clear historical record. But --precisely because I
continue to be interested in the interactions between process, formal
procedures, and results-- I don't think it is productive in this
case, especially to what, IMO, is the extreme to which you want to
take things. More specifically, I would personally prefer to avoid
turning the IETF, or any of the related or subsidiary structures,
into an experiment in political theory or utopianism (mine or anyone

With regard to your substantive comments about appeals and jury
processes... The reality is that all of the appeals that have reached
the IAB within my memory have involved significant and often complex
technical components, typically including questions of whether WG, WG
Chair, or IESG decisions on a particular issue were reasonable in the
light of some technical argument. They haven't been exclusively
procedural, or exclusively technical, but a inseparable mix of the
two. In the "outside world", the historical experience with
randomly-chosen juries in that type of situation has been fairly
terrible. That is not surprising: if one doesn't understand the
issues but is forced to make a decision, one decides on some other
basis. It is also well-known that, if the randomly-chosen members of
the jury can be educated about the issues, then results get better:
but that is a time-consuming process and can easily lead to claims
about the biases of the educators who must, recursively, be selected
and/or monitored by some process that reduces or eliminates those

Now, if we were dealing with exclusively procedural appeals, I would
agree with you that an independently-selected appeals body, one that
could deal with conflicting claims about what had occurred on an
adversarial and evidentary basis, would be a better and more
convenient idea. I actually suggested that a few years ago, possibly
in the document from which you extracted the above, because I assumed
we would see mostly that type of appeal. But, if I recall, the
community rather throughly rejected that idea. And, in retrospect,
I'd assume that at least part of the reason was that I was wrong:
while there would still be some attraction to getting appeal work off
the IAB's agenda (my other motivation at the time), we haven't seen
those types of appeals.

I could quibble with some of your other suggestions (e.g., picking
people conditioned on attendance would result in an uncomfortably
biased (in the statistical sense) group, especially if an appeal
addressed the way in which a mailing-list-only participant was being
treated. But it isn't worth it: the general idea itself is fatally
flawed by an incorrect assumption about the nature of the appeals
that arise.

Conversely, blue-ribbon appeals panels for more technical issue have
some logical appeal, but would, I believe, create a different sort of
bias, one that involves very slow procedures and responses. And a
non-trivial fraction of IETF participants seem to have a deep
suspicion that people who _really_ want to serve on such things
should be disqualified as non-representative of the community, if not
outright insane, which is probably not a good basis for trust in such

The system isn't perfect, but options such as the ones you suggest
were considered and the conclusion was that, on balance, the present
setup isn't significantly worse than other possible choices and
better than many of them. And I think actual experience so far, as
distinct from theorizing about possibilties for what might happen or
about ideal processes, has largely confirmed the acceptability of
those choices.


p.s. The above remarks have not been examined to be sure that no word
could possibly be misinterpreted by someone inclined to do so. So
now I guess I get to spend the next year or four worrying about which
phrases will be taken out of context and used ot demonstrate
something I don't believe (either now or "by then") :-(

This message was passed through, which
is a sublist of Not all messages are passed.
Decisions on what to pass are made solely by Harald Alvestrand.

Date view Thread view Subject view Author view

This archive was generated by hypermail 2b29 : Sat Feb 12 2000 - 09:24:51 PST