IETF 61 Plenary Meeting Minutes
IETF 61 Plenary Meeting Minutes
For the first time in a while, we held only one plenary meeting at IETF
61.
Hosts: Harald Alvestrad <harald@alvestrand.no> and Leslie Daigle
<leslie@thinkingcat.com>.
Minutes: These minutes were taken by
Spencer Dawkins <spencer@mcsr-labs.org> and cross-checked with
the session jabber
logs (available at
http://www.xmpp.org/ietf-logs/plenary@ietf.xmpp.org/). Thanks
especially to George Michaelson <ggm at dstc.edu.au> and Suresh
Krishnan <suresh.krishnan at ericsson.ca), who type like madmen, and
Scott Brim, Donald Eastlake, and Mark Thompson, who provided
corrections on a first draft posted to the IETF mailing list.
Welcome, and
introduction - Harald Alvestrand
- 1314 attendees this IETF, pretty close to Seoul and San Diego
- Number of countries is down from 40s to mid-20s
- Still getting work done
- 4 new WGs, 11 closed working groups
- lots of IDs (especially in last couple of weeks, as usual)
- Looks like we're headed for a new world's record in publishing
RFCs this year...
- Spring IETF in Minneapolis (March 6-11), Summer in Paris, France
(France Telecom sponsoring), working on Fall location in Canada next
year, no snoeshoes will be required (darn!)
Host's presentation - Steve Demetrio
This presentation was actually quite funny - I'm sorry the notes and
jabber logs only captured part of the humor
- When traveling to IETF, we found out that even fingerprinting
technology is incompatible ...
- Why don't the Host RFCs actually help IETF hosts? At least none
are written by Al Gore...
- Gee, we've always been light on presentation technology (for more
than 35 years!)
- I remember all the events of my 14th year except the birth of the
Internet...
- We owed it to the IETF to play our part as hosts for this event,
and we learned a lot
- Recognition of Susan Brewer and Ben Crosby for special efforts in
making this event happen
A word from the NomCom - Danny McPherson
- Looking for feedback, guidelines in RFC 3777, until 11/23/2004
- nomcom04@ietf.com, or talk to someone with an orange dot
- "retaliatory nominations not considered!"
State of operations:
RFC Editor report (Aaron Falk)
- Document arrivals steadily increasing, professional copy- editor
now in use - authors may see more queries now
- RFC Ed desk hours at IETF seem to be popular - plan to continue
at future IETFs
- New Internet Draft IPR boilerplate is available (new RFC numbers
are 3907/3908, but recommend referring to these as BCP
78/79)
- Getting more XML than we used to - please send us your XML if you
use it
- Joe Touch has put together a new MS-Word template - feedback to
Joe or to us
- Submissions and publishing are running neck-and-neck
- Fred Baker - I thought all RFC-Editor inputs were unsolicited.
Having a spam problem (groan!)
IANA report (Doug Barton)
- Spending time looking at draft queue dwell time - coming very
close to goal of "less than four weeks" now
- Have hired new staff (Pearl Liang), and Michelle is working with
IETF more closely
- Want to be more transparent
- Queue is about 20 documents now, average age is decreasing to
about 4 weeks (thank you, Michelle!)
- Meeting with IESG and IAB to improve reporting on performance and
goals
- Queue status page is now online
- Private Enterprise Number registry and port registry requests
current or close to it - allocated over 300 PEN numbers last month
IESG operations (Allison Mankin, Bill Fenner)
- Training copy editors is really important...
- Presenting update of same data described in San Diego
- We're approving more documents than requested - we're getting
ahead
- Numbers are lower, but interval is 14 weeks against 22
Seoul-to-San Diego
- Will continue to make this data available and to evaluate
upcoming PROTO experiments - we're interested in suggestions
Review of Architectural
Activities
IAB Chair report (Leslie Daigle)
- Seeking inputs on Link-Layer Indications and Liaison
Relationships IAB drafts (see our web page for pointers)
- OMA-IETF Liaison Report is approved and in publication queue
- Meeting with ISOC on Administrative Restructuring
- Pete Resnick reported on Messaging Workshop on consent-based
messaging, supporting IRTF, IESG, and IAB
- Solicited write-ups and did a two-day workshop of breakouts
- Thought about key distribution and identity, ignored canonical
ratholes ("elephant in the middle of the room")
- Worked on engineering, research, and architectural topics like
automated introductions, consent-based frameworks (similar to IM)
- Will publish report and begin thinking about future workshops
(this was first workshop in two and a half years - high startup
transient, but people came, worked hard, not smooth but came out of the
workshop talking)
Administrative
Restructuring (Harald and Leslie)
- (As Harald and Leslie started this section, they announced that
Steve Bellovin is stepping down as Security AD, and is being replaced
by Sam Hartmann)
- We want to present where we are and where we are going and talk
about the AdminRest draft
- Does NOT affect the IETF standards process.
- We believe there is community consensus for moving forward Admin
as an ISOC activity - defining documents are still drafts ("speak
now"), and we're starting to work based on these drafts
- We've been working on this since August 2003, and pretty intensely
- Have joint IAB/IESG draft on admin restructuring
- Have discussed with ISOC board
- Goals
- Keep IETF going in 2005
- Get BCP approval for what we're doing
- Make decisions by people responsible for long-term success
- Want to move quickly, be transparent, be effective
- Have added editor team for BCP drafts, hoping for more community
discussion
- Will ask NOMCOM to fill IAOC slots when BCP approved (four week
last call)
- Would love to approve at January 6 telechat
- Want to appoint a transition team for restructuring this month -
looking for good candidates, please send e-mail to IAB/IESG
- Want to seat initial IAOC in January/February timeframe, with
contracts in place by mid-2005
- None of this should be news, right?
- Won't get to RFP before 2005, CNRI is working on new proposals
for secretariat in 2005 - IASA transition team will review, may be able
to sole-source for a period of time and then integrated with IASA
- Harald presented document content and open issues
- Read the documents - devil IS in the details.
- IASA continues close relationship with ISOC
- Continued separation between ISOC fundraising and IETF
standards process
- ISOC still has oversight
- Remember- no change to IETF technical process
- The acronyms are endless (this is an IETF activity, so...)
- "One, single, salaried person who makes sure we have
administrative contracts that are executed and meet IETF needs"
- Almost everything is outsourced...
- Seven voting members (two chosen by IETF NOMCOM, one by IESG,
one by IAB)
- Two funding sources (unchanged from today, but go into one pot)
- Budget determined each year (details in the draft)
- The BCP is an IETF document and will be last called on IETF
list to the community.
- ISOC must approve this, too (since they have to do their part)
- Need to set timelines for the dates we're shooting for
- Issues remaining
- Final composition of IAOC?
- Questions?
- Dave Nelson - is new admin process more or less expensive than
previously? We're adding one paid person, don't know about the total
yet. Need to move overall administration under one umbrella - hope for
efficiencies, don't know for a fact. Best-guess 2005 budget posted to
IETF list this morning.
- Michael Richardson - will we get better organization? Will we
know where the meetings are? If not, it's our fault.
- John "Trouble-maker" Klensin (his words!) - two observations
and a question - can you put
out critical documents that aren't on top of plenary meetings? We are
making progress (documents were out two weeks ago). Not a complaint,
just a request. The timeline is still a draft. Timelines have shifted -
why? Will they shift in the future? We're adding external review by
experts - great! Where does it fit in the timeline? Still thinking
about this, because we just added external review yesterday morning.
Keep an eye on new versions of the BCP draft for details, and guidance
on advisors is appreciated. The timeline has been out since 00 of the
Internet Draft, to be fair.
- Richard Perlman - based on budgeting proposal this morning,
much of our funding will come from public registry income - how
dependent will we be on an external organization? We had multiple ideas
on where to go in the consultant's report at San Diego - we think this
is the lowest-risk option. Go home and tell your organizations to
contribute to ISOC! Lynn - it's actually $840K, and next year is $1.4M,
with a
significant increase (RFC Editor, for example). This is not an
aggressive plan, not a conservative plan. Thinking about a long-term
commitment that accommodates other ISOC activities (education, etc.). -
Ted Hardie - look for details about operational reserves in the draft,
what it covers, etc. Recognize that we need to raise additional funds
in some way.
- Brian Carpenter - we can evaluate the risk when we see a budget
- thank you for sending it. This is already an improvement - we don't
know what today's risk actually is! Should the IAB chair be voting or
ex officio? My preference is voting. - Margaret Wasserman: all the
proposals had different membership proposals, and we left them
different to stimulate discussion. My preference is actually freeing up
IAB to do architecture and not worry so much about stuff like this.
Brian - this is now another external liaison, of course... Community
discuss, please!
- Bob Kahn - I starting out thinking restructuring is a bad idea.
I've come around. Some degree of restructuring is appropriate based on
organizational maturity. I (CNRI) am optimistic, but do have some
concerns: Any new restructuring has to stand on its own and meet real
needs. There are needs that haven't been considered. Please look at
Patrice's proposal (it's a current draft). How to define performance
levels for support organizations and see if they are being met? The
IETF will have additional needs in the future - need to think about
this, too. The Internet has the attention of everyone in the world -
the world needs to be comfortable with the new structure, too, and
recognize that not all the issues will be technical ("public interest
at large"). And don't count on sharing resources - the existing support
structures are all under stress! Go with lightweight structure while we
make the transition.
- Harald - recognize that the IETF chair wasn't picked for his
expertise in contract management... We have been aggressively
open, and this has been our
defense against charges of conspiracy. We need to stay open and stay
relevant.
- Dave Nelson - this may be necessary but not sufficient for
long-term survival. What about long-term economics? Our work seems to
be getting more expensive as time passes. San Diego presentation about
all the organizations that want to play a part in "the Internet".
Are we concentrating at the right place? Harald - this isn't actually
part of AdminRest, but it does matter
(Leslie agrees here). We need to be financially transparent and
responsible.
- Aaron Falk - RFC editor has added staff, has increased our
staff cost, don't know about effect on overall costs. I sympathize with
the work you guys have been doing (speaking as a liaison participant).
Are there already new BCP editors? What do we do with our feedback?
Harald - we've gotten suggestions for editors inside and outside the
IESG. Rob Austein and Bert Wijnen will be editing the next draft. Send
feedback to IESG or IAB, but issues to IETF list - we
haven't gotten a lot of feedback yet.
- Spencer Dawkins - I've been watching the IAB and IESG trying to
balance technical work and admin work for a while now, and I know it
hasn't been easy - thank you! (applause for the IAB and IESG)
- Patrice Lyons - a lot of organizations at this level of
maturity formalize their status (it's BEEN 35 years). Should you have a
formal existence? Should you have an Executive Director that reports
directly to the organization, not to some other group? Third-parties
providing legal advice aren't tied to this organization! Should the
IETF determine its own budget? IAB/IAOC/ISOC isn't the IETF, or even
the IETF leadership. Leslie - there's a lot more detail in the draft
than on the slides - this really is an IETF budget. Maybe legal counsel
really is required for some of this - we've had external counsel before.
- Ted Hardie - at some level, the IETF is "a collective delusion
with brownies at breaks" -
the IETF is so anti-organizational! becoming a membership organization
probably makes sense, but the feedback we got was "no". We're trying to
do the least we can do that allows the community to continue to do the
volunteer work we've done in the past. If we don't listen to the
community, they should fire us.
- Patrice - speaking as counsel to CNRI, but I'm a member of the
community, too. You're talking about million-dollar budgets, but you're
relying on third parties to plan your future. That's not what you want.
Take six months, read my draft, think it over.
- Harald - we decided not to incorporate in 1994, pretty
strongly, and I'm still hearing support for this decision now ("we're
fine as we are"). We just haven't seen any sign of community consensus
to change this.
- John Klensin - I've been concerned for six months that we would
make imporant decisions (potentially fatal choices) without adequate
community input. We're doing a lot of good things, but I'm concerned
that (for instance) we may hire our "one person" according to the wrong
job description. Also - there's a rumor about internal organizational
changes at the secretariat - would anything like this disrupt the
process or blindside the community? Also - IANAL, but I've hung around
other standards bodies. In the US, there's an American National
Standards Institute - they don't create standards, they accredit
standards organizations - the vast majority of accredited organizations
are not separately incorporated - doesn't agree with what Patrice was
saying about mature organizations.
- Bill Manning - I've read these documents - I worry about
attrition, about people voting with their feet if the IETF becomes too
onerous. Are we losing return attendees?
- Tim Shepherd - What is the IETF? In some sense, it's only the
administrative support organization, because we can vote with our feet!
We're all volunteers. I'm not worried.
- Patrice Lyons - Speaking as the CNRI counsel - we're not just
moving papers around, we're
talking about funding, and funding can bring ties if you're not
careful. My proposal didn't require incorporation, only formal
existence - the NYSE ran that way for years.
- Dave Nelson - people aren't walking into the IETF, we haven't
had any RFCs for BitTorrent, for E-Mule, for KAZAA. Are we moving too
fast?
- David Black - I spend time in ANSI organizations, and John is
slightly wrong - IEEE is incorporated.
- Spencer Dawkins - in 1996, we were swamped, and in 2000 we were
encouraging people to take work elsewhere (especially in APPS and
Sub-IP). In 2004, is anyone thinking about increasing IETF
participation? Harald doesn't want to talk about this as an issue of
restructuring.
- Harald - how many people think we're moving too fast? too slow?
about right? Allison - how many people want us to just go away?
- Bob Kahn - I'm hoping at the next meeting I'll be able to
present proposals for CNRI changes, but, in the meantime - we never
wanted to be in our role forever. We want to shed our "public interest"
role - don't know how to offload that yet. We also want to develop a
mechanism for operational change control - we're heading there. Things
are costing less - that's a problem! Attendance is down, support is
down - and that's not good. We want to turn a lot of CNRI oversight
responsibilities directly to the IETF. Foretec may be a player in this,
going forward, but we need to structure this relationship
appropriately. About consensus - the Tao of the IETF for decades has
been running code and rough consensus. This plan isn't running code yet
- how can you have informed consensus?
Joint IAB/IESG Open
Plenary
- Bill Manning - IETF should no longer work on IPv6. We should work
on IP or replacements of IP. Calling out versions leads to
divisiveness.
- David Perkins - I was recently SPAMmed by IPv6 Summit because I
came to the IETF. Do we have a zero tolerance policy against spammers?
People who harvest contact information? Require double-opt-in? Harald -
we're still working on preventing other people from spamming using our
systems! But it sounds like a plan - send a draft.
- Tony Hain - I agree with Bill Manning - remove version numbers
from working group names. We're stuck, and we need to move forward, or
we'll be roadkill.
- Margaret Wasserman - we're not going to transition to IPv6 the
way we
expected - we need to focus on a network that has both, for a very long
time.
- Donald Eastlake - main reason attendance is down is because of
the dot-bomb crash as the high-tech bubble burst. People have been
predicting the death of the IETF for a
while.
- Fred Baker - on SIEVE (RFC3028)/whitelist capabilities - there's
a BoF
tomorrow on policy-driven white-listing, and we should apply this to
our own mailing lists!
- Spencer Dawkins - concerned about weird devices that transform
packets in
the network - BEHAVE, SIPPING ad hoc last night on Session Border
Controllers. What if the wasp waist of the internet stops being IP?
That's been the threat for ten years, and HTTP isn't the alternative
like it was in the mid-1990s. Jonathan Rosenberg - share your concern,
we need to meet this head-on and deal with it. Spencer - not
criticizing our response and appreciate what we're doing now - I'm just
concerned. When can we think about architecture some more?
- (sorry - missed name, but someone from IIJ) - concern is IPR.
Interested in a protocol,
but it has IPR. Harald - IPR is tricky - we're trying to do what we
can. Steve Bellovin - community hasn't changed enough to change our
stance on IPR since the last time we thought about it, and we can't fix
the problem as a whole from inside the IETF.
- Nico Williams - Security BoF at last
IETF - it's not about security, it's about contracts. Can more be done
to help do business on the Internet? Russ - we expect to see a repeat
BoF at Minneapolis.
- Harald - we need more informed consent on AdminRest. We have one
Internet, not two, and we need to act like it.
|