Re: IAOC: delegating ex-officio responsibility
Speaking as an individual, but an individual who helped set up this
structure and who sat in the non-delegated ex officio IAB Chair position
on the IAOC (and IETF Trust) for a couple of years, let me offer some
As Bob noted, elsewhere in the thread, your draft does not describe how
to delegate responsibilities, it describes how to allow alternatively
fill the positions. And I believe that is a fatal flaw.
The discussion on this thread certainly indicates that there isn't a
clear grasp of what the responsibilities are for the positions in
question. The responsibilities are not: show up for meetings/telecons,
offer opinions about how to administer the IETF, and put in a vote when
called upon, although those certainly are the functions expected.
You described the IAB Chair as "hub" as if that was a bad thing. In
fact, the one thing the Chair (of the IAB -- but it applies in analog to
the IETF) needs to do is to support the IAB's functioning (through
leadership, organization, note-taking, cajoling, listening, etc). That
doesn't mean the IAB Chair has to do the work of the whole IAB, but it
does mean that an overall perspective is critical, and each Chair has to
best organize her/himself to achieve the one task.
In setting up the IAOC as it is, the intent was to inject the IAB
Chair's overall perspective into the IAOC's discussions, including the
support for IAB activities. It was not a question of simply finding
more credible people to put on the IAOC.
If change is needed, the one path away from the fatal flaw seems to be
to articulate the actual responsibilities of the IAB Chair (and IETF
Chair and ISOC CEO, as appropriate) in the intended ex officio
positions, such that it is possible to get agreement and external
confirmation that the role is being properly fulfilled by whomever is
selected to perform them.
Anything else, IMO, is mindless deconstruction.
On 4/14/11 5:01 AM, Olaf Kolkman wrote:
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
On Apr 14, 2011, at 9:17 AM, Brian E Carpenter wrote:
On 2011-04-14 06:19, Bob Hinden wrote:
On Apr 2, 2011, at 1:28 AM, Olaf Kolkman wrote:
It seems that the high order bit of this discussion circles
around the question on whether it a requirement for the IETF
Chair to have a voting position in order to effectively perform
oversight. Once we figured out where we want to go with that we
can think about delegation by the chair vs appointment by the
bodies and the implementation details with respect to the
For the record, I don't agree with this summary. That is, I
still question the basic assumption in the proposal. We have
"running code" in the IASA model and it appears to work
reasonable well. Not perfect, of course. In particular I think
that having the IETF chair, IAB chair, and ISOC president as
voting members of the IAOC (and IETF Trust) has worked very well.
It makes them an active part of decisions the IAOC and IETF Trust
are making and helps keep the IAOC from getting disconnected from
the community. It also makes them share the responsibility for
decisions by having their vote be publicly recorded.
I agree. I think this responsibility should not be delegated. It's
fundamental to the success of the IETF (and the ISOC for that
matter - the IETF is a major source of ISOC's legitimacy). The IAOC
delegates execution to the IAD etc. - maybe the real bug is that
the IAOC itself needs to delegate more?
I agree the IASA model is working well and I also agree that the
chairs have played an important role.
But I am not convinced that it is the only way to keep the IAOC from
getting disconnected from the community. To turn that argument
around: If it is the case that we need the IAB chair, IETF chair and
ISOC president [iChiefs] to keep the IAOC on track then why do we
have NOMCOM appointments? One could easily argue that the NOMCOM
appointments are made to keep the IAOC connected to the community.
I also do not see why it 'it's fundamental to the success of the
IETF'. I do agree that the iChiefs need to understand what is going
on, there is no dispute about, but I question whether IAOC membership
is best and most efficient instrument.
I also don't understand what the effect of the proposal is on the
IETF Trust. Currently all IAOC members are members of the IETF
Trust. They have to sign a letter accepting this role. I don't
think it can then be delegated.
It can't be delegated. However, a duly approved update to BCP101
can change the definition of the formal membership of the IAOC, and
that would automatically change the membership of the Trust. An
alternative would be a formal change to the Trust Agreement, but
since IANAL, I don't know whether a change to allow delegation or
proxies would be possible under the law of the Commonwealth of
Virginia, where the Trust was established.
I do not want to brush over this important detail. But, let us first
figure out what we want with the IAOC. If there is no consensus on
having the iChiefs play a less involved role then fine tuning the
details doesn't make sense. If we gain consensus then fixing the
Trust is doable.
Your draft focuses on one area (that is, reducing the burden of
these positions), but does not discuss any other aspects of
making this change. What might the negative aspects of
delegating this responsibility be? How will this be dealt with?
Help me out.
The main _potential_ negative aspect is a drop of iChiefs' awareness
of what is going on; and the ability to directly influence
Each of the positions (IETF Chair, IAB chair, ISOC President) are
different in the way they are selected and this effects their
ability to delegate their responsibility and who they might
delegate it to. For example, the IETF chair is selected by the
NOMCOM and one of his/her responsibilities is to sit on the IAB,
IAOC, and IETF Trust. The IAB chair is selected by the IAB. The
ISOC president is hired by the ISOC Board of Trustees.
Consequently, I think the authority to delegate differs and they
should be considered separately.
I agree, this needs discussion. But lets first see if there is any
consensus on going down the path of getting the iChiefs out
day-to-day IAOC business.
I agree that the IETF chair needs to have a good oversight
about what goes on in the IETF, to a lesser extend it is good
that the IAB has that oversight too (specifically with respect
to its chartered responsibilities) but I wonder if a voting
membership is the appropriate instrument.
Why not? It does appear to work.
Yes it works. But it consumes a lot of time too. IAOC retreat,
meetings at the IETF, regular tele-chats, participation in sub
committees, etc, etc...
We have a responsibility to our leadership to make their jobs
scalable and manageable. We need to have our leadership positions
accessible to our peers, who may not have the support of their
company for more than a half time job, or, not to forget, may not
have the support of their families for the travel and the long
evenings with telechats. Being effective in what committees and
meetings one participates in is part of making the job scalable.
The Best Current Practice is that the iChiefs _must_ participate in
the IAOC, I like that _must_ being removed.
I believe effective oversight depends on having the appropriate
high level information and having the opportunity to timely
inject information that is needed to steer an outcome. An
alternative method for sharing and injecting is having regular
meetings between the I* chairs and the ISOC President/CEO. I
believe that such meetings are much more effective for the
parties involved than being exposed to all details.
Do we really need to have another regular meeting? Would this
give the I* chairs more authority than they have now? Sort of an
executive committee. Would these meetings be public, have votes,
have public minutes?
It is not another meeting, it is replacing a lot of other meetings by
one regular meeting that talks about issues at the right level of
detail. Also, sharing information is not decision making. Russ and I
had weekly calls planned (we didn't have them weekly, but that aside)
to share information and brainstorm about potential issues.
I don't see the 'more authority' argument, not in the context where
one argues for direct votes in order to get effective oversight. I
think that having a weekly iChiefs meeting gives less formal
authority and (yes) a bit more informal authority.
This only an illustration of an instrument, there may be other
instruments for oversight as well. But I do not think the
ex-officio membership is the only method.
It's not perfect but it is the one we have now and it is working.
We should only change it if we are sure that it will improve the
overall IASA operation. I am seriously concerned about the
My question is whether there is any will to pursue a change to the
BCP to work towards removing the _must_. (The Best Current Practice
is that the iChiefs _must_ participate in the IAOC, I like that
_must_ being removed.)
I have seen Brian and Bob argue against. And not very many people
Olaf M. Kolkman NLnet Labs
http://www.nlnetlabs.nl/ I will start to use a new PGP key (ID
0x3B6AAA64) at the beginning of May 2011.
-----BEGIN PGP SIGNATURE----- Version: GnuPG/MacGPG2 v2.0.17
(Darwin) Comment: This message is locally signed. Comment: GPGTools -
cH4AoI0vilyFYRIa5Bx1j6i7T45klrYs =F6bJ -----END PGP SIGNATURE-----
_______________________________________________ Ietf mailing list
Yours to discover."
Ietf mailing list