On 2 sep 2004, at 07.11, Brian E Carpenter wrote:
Yes that would be helpful.
Well, I don't agree. I think it would defocus the discussion (which
is about putting the IETF's administration onto a business-like
basis). IMHO the only case in which we should discuss the wider
option is if the newtrk WG proposes changes in the standards process
that would make such a thing necessary.
I guess I don't understand this comment.
As I see it, one reason for another option would have to do with the
independence of the IETF to change its processes, should it want to.
Not necessarily because it has a plan today to do so.
I see no respect in which the creation of a new non-profit organization,
which for the sake of argument we can call ISOC2, would affect this.
ISOC was to a very large extent created for the benefit of the IETF (and
created by people quite heavily involved in the IETF at that time).
Apart from its direct support and legal umbrella function for the IETF,
it also offloaded from the IETF a lot of "outreach" functions (otherwise
known as Layers 8 and 9). We mustn't forget that if ISOC hadn't existed,
the IETF would have been much more distracted by Layers 8 and 9 over
the last ten years.
One of the concerns I have over the ISOC dependent mechanisms, which I
guess is all of the presented options, is the link between budgeting and
process. If a process change requires a different form of budgeting
support, would the IETF need the approval of ISOC to make that change?
Not if we establish the appropriate MOU and (I would suggest to ISOC)
make appropriate amendments to the ISOC by-laws to guarantee the
Often what seems like a purely technical decision has policy and
Absolutely. That is the fundamental reason we need clarity in all
Assuming that we don't want to have reconsider
the organizational relationship again in the near future, I believe we
need to take such possibilities into account.
Absolutetly. That's why in any case, we need the MOU and by-laws that
I mentioned -and we need them whichever scenario is adopted.
I think another consideration in making these administrative decisions
has to do with the IETF's voice in the general standards and Internet
governance arena. Will ISOC, as a 'parent' organization - my
interpretation of the options that are offered, be the responsible party
for such activities?
It is today. That's why it has a VP for Public Policy. [Disclosure: that
person happens to be my day job manager.] Are you (not just Avri) an
active ISOC member? It's free...
E.g. currently for a liaison to the ITU, it is
ISOC that is the liaison association. Should ISOC disagree with the
IETF position on a liaison matter who has the final say?
Good question, and a point not to be forgotten in the MOU, in every
the ongoing governance debate in the international arena, will ISOC or
the IETF be the negotiating body?
Today, it's ISOC. ISOC has been working to position itself in that
debate for years; having ISOC's European office in Geneva was not a
random choice. Put bluntly, if we threw that away and started again
with ISOC2, it would take 5 to 10 years to build up the momentum again.
I just don't see why we would want to do that.
And before we decide that this is
just policy and does not relate to protocol issues, we should not ignore
the intimate link between policy and technical - while it is not always
direct, there generally is a technical implication in policy decisions
and, generally, also a policy impact in technical decisions.
Indeed. A good example was the recent ITU/WSIS meeting on spam -
John Levine attended and spoke for the IRTF, funded as Harald mentioned
by his budget from ISOC; I attended (because I live locally) wearing
an ISOC hat for the occasion. ISOC's reputation got us in, but
the representation and impact was basically IRTF/IETF. It works today,
with ISOC1. I just don't see why it would work better with ISOC2.
I am concerned about the real independence of the IETF as a technical
standards body when ISOC, which the IETF does not control, has the
governing policy and financial voice.
The IETF, as a bunch of people who choose to meet and send email,
will have that concern with *any* legal entity that acts for it.
It all depends on the rules of that entity, not on the desires
of the IETF population - the trick is to make those rules
right, and I deeply believe that can be done soonest and cheapest
by recycling ISOC.
I would be interested in seeing an analysis of an option which has the
IETF as a independent nonprofit corporate entity. This could be either
as a wholly owned subsidiary of ISOC, thus keeping the fiduciary
relationship, or as completely independent organization.
The subsidiary model was actually taken out of Carl's scenarios
before publication, but apart from that I thought this was exactly
what scenarios C and D are.
The objection I heard to the subsidiary model was that it really
couldn't give the IETF community any more guarantee of control
than scenarios A and B, but would create the overhead costs of
scenarios C and D anyway.
Ietf mailing list