IETF 76 Administrative Plenary
Wednesday, November 11, 2009
Minutes by Karen O'Donoghue
Agenda
======
1. Welcome
2. Host Presentation (with RFID Report)
3. Itojun Service Award
4. Reporting
- IETF Chair Report
- NOC Report
- IAOC Chair & IAD Reports
- Trust Chair Report
- NomCom Chair Report
5. Recognition
6. IAOC Open Mic
7. IESG Open Mic
Welcome, Russ Housley
=====================
Slides: plenaryw-1
plenaryw-8
Russ Housley called the IETF Administrative Plenary to order. He
presented the agenda and the IETF Note Well slide.
Host Presentation, Jun Murai
============================
Slides: plenaryw-7
Jun Murai from the WIDE project gave the Host Presentation. He
acknowledged the many people that made the meeting possible including
the administration of the City of Hiroshima, the mayor, and the
T-shirt designer. He noted the many ways in which Hiroshima went out
of their way to welcome the IETF.
The remainder of his time was spent on the RFID experiment conducted
at this IETF meeting. In general, the technology was more widely
utilized than expected. There were two types of RFID used (ISO 15693
and ISO 18092 subset). Privacy concerns were addressed in a number
of ways including an Opt-in/Opt-out structure. A great deal was
learned during the execution of this experiment.
Russ Housley thanked Jun Murai for his efforts hosting this very
successful meeting. The plaque thanking him for his efforts will be
shipped at a later date.
Itojun Service Award, Jun Murai
===============================
Slides: plenaryw-0
Jun Murai introduced the Itojun Service Award established to honor
the memory Itojun and his contributions. The first Itojun Service
Award is to be presented is being awarded in Japan, and Itojun's family
has been invited to help make the presentation. Itojun's mother and
brother were introduced. Itojun's mother addressed the group
acknowledging Itojun's passion for IPv6 and the friendships he had
in this community. Jun Murai presented the award to Lorenzo Colitti
and Erik Kline from Google for their sustained efforts to make
significant commercial contents available over IPv6. Lorenzo and
Erik recognized additional contributors within Google and those
outside Google, including Itojun, who made their award possible.
Reporting
=========
IETF Chair Report, Russ Housley
-------------------------------
Slides: plenaryw-4
Russ Housley provided the IETF Chair Report. This report included
the IETF 76 attendance numbers and the IETF, RFC Editor, and IANA
activity since IETF 75. Russ also acknowledged the volunteer
activities including the WebEX experiment (Joel Jaeggli) and the code
sprint (Robert Sparks, Ben Campbell, Pasi Eronen, Henrik Levkowetz,
Arifumi Matsumoto, Lars Eggert, Tero Kivinen, Tony Hansen, and Adam
Roach). Russ closed the report with the list of future meetings.
NOC Report, Akira Kato
----------------------
Slides: plenaryw-6
Akira Kato from Keio University and the WIDE Project presented the
NOC report including technical details of the network and
acknowledgements of the companies and individuals that made the
network possible.
IAOC and IAD Chair Reports, Bob Hinden and Ray Pelletier
--------------------------------------------------------
Slides: plenaryw-5
Bob Hinden announced that IETF 79 will be in Beijing. He also pointed
out that the IAOC is working to improve transparency and is focused
on listening to the community. Next, Bob provided a financial update.
There were significant concerns going into 2009 about what the year
would hold financially. However, a number of the steps taken early
on have resulted in a reasonable financial status at the end of this
year. Bob also provided the 2010 budget and projections. More details
are available in the slides.
Next, Bob discussed the ongoing effort to restructure the RFC Editor
into a new model with four components: 1) RFC Production Center, 2)
RFC Publisher, 3) RFC Series Editor, and 4) Independent Submissions
Editor. The RFC Production Center and Publisher contracts have been
awarded to AMS and the transition is underway. The RFC Series Editor
and Independent Submissions Editor portions are still being
addressed by the IAB.
All the existing contract cycles were summarized and a short
discussion on additional contract work being pursued was provided.
The currently scheduled future meetings were summarized and the host
opportunities identified. Finally, acknowledgements were provided
to the Secretariat, RFC Editor, IANA, NOC volunteers, WIDE, and the
many sponsors that made this meeting possible.
Trust Chair Report, Marshall Eubanks
------------------------------------
Slides: plenaryw-10
Marshall Eubanks provided the IETF Trust Chair Report covering IETF
Trust Actions since IETF 75. The first topic was issues with and
status of the modifications to the Trust Legal Provisions (TLP). This
has been a significant effort. In response to community concerns,
the IETF Trust developed and adopted a standard procedure for
modifying the TLP. In addition, a great deal of work was put into
the issue of non-IETF streams. Also, the IETF has received one
subpoena since IETF 75. Finally, the IETF Trust is working to
improve communication with the community.
NomCom Chair Report, Mary Barnes
--------------------------------
Slides: plenaryw-3
Mary Barnes provided a report from the current NomCom process. The
nomination period has closed, and the NomCom is currently in the
process of gathering input. RFC 5680 "openlist" requirement does not
apply to this NomCom period. IETF participants are strongly
encouraged to provide input to the NomCom.
ISI Recognition, Olaf Kolkman and Russ Housley
==============================================
Slides: plenaryw-9
Olaf and Russ presented thank you plaques and glass bowls to Bob Braden,
Sandy Ginoza, and Alice Hagens from ISI for 30 years service to the
IETF. There was a presentation and reception in their honor earlier in
the week.
IAOC Open Mic
=============
Bob Hinden called the IAOC Open Mic into session. All the current
members of the IAOC came to the platform and introduced themselves.
Tim Shepherd brought up the issue of the as yet unpublished RFC 5569
that is stuck in Auth48. He doesn't understand why we couldn't have
done something faster. He wants to ensure that the whole IETF
understands what a problem this has been this year. Marshall Eubanks
responded by stating that unfortunately it is going to take a bit more
time because of the TLP issue. The whole independent submission stream
of documents is on hold while the issues associated with the TLP get
resolved. Jari Arkko indicated that he is partially to blame for the
delay in Remi's document. Can we agree that the next time we change the
boilerplate, we leave the old one in place until the new one is ready?
Remi Despres pointed out that the document describes how the provider
deployed IPv6 in 5 weeks, and that it has taken more than 5 months
to get the document approved.
Aaron Falk pointed out that we found a more heavy weight process than
publishing RFCs for managing our process, but that you can actually get
to the information in the approved-for-publication Internet-Draft.
Fred Baker reminded folks that there are a number of people who will
wait until something is in archival state until they use it.
Tim Shepherd pointed out the problem with referencing the Internet-Draft
before it has been published as an RFC. The fundamental mistake we made
in this case was not synchronizing the process so that the update to the
rules for the independent submission stream and the other streams are
synchronized.
Another gentleman appreciated that concerns about free speech were
acknowledged, but he felt that they had not been seriously addressed.
Bob Hinden indicated that we got a proposed hotel contract that we
weren't comfortable with, asked for community feedback, asked the
host to talk to the hotel because we didn't feel we could proceed
with the clause, and the host got the hotel to remove it.
Spencer Dawkins asked if more could be said on where we might be headed
with WebEx support in the future. Bob Hinden responded that we've
started doing WebEx on a limited basis. Russ Housley asked for
feedback on the experiment. There is some discussion of making WebEx
available to working group chairs to have virtual interim meetings
as well. Joel Jaeggli, as the key volunteer supporting the experiment,
witnessed a situation where there were two contributors, one on jabber
and one on WebEx. The conversation was being moderated by the working
group chair. This doesn't seem like a situation we want to encourage.
Chris Liljenstolpe asked if we will have unblocked Internet access
in China. Bob Hinden responded that this is specified in the MOU which
references the IETF network requirements.
Douglas Otis had a hard time in Stockholm because of display issues.
He suggested that we adopt a strategy to provide more flexibility
and capability where displays are concerned.
Gregory Lebovitz asked if the network will have helpful network
analysis tools. The standard network requirements that we have used
elsewhere are included in the Beijing MOU. Joel Jaeggli indicated
that this would be an excellent application of DNSSEC validation.
Pete Resnick pointed out that we have three consecutive years with
an Asia meeting. He thought it was supposed to be 50% North America.
Ray Pelletier responded that the IAOC adopted a 3-2-1 cycle for every
two years, with 3 meetings in North America, 2 meetings in Europe, and
one meeting in Asia. A change took place in 2010 because of feedback
on the US visa issue. It was decided to defer Atlanta and backfill
with another location. The IAOC has not adopted a change in the
policy. Jari Arkko pointed out that according to his tools, 64% of
Internet-Drafts come from North America and 29% from Asia. It might
be time to reevaluate that policy.
Lou Berger asked about the rationale for back-to-back Asia meetings.
Ray Pelletier let everyone know that the IAOC had already asked him to
work on changing that situation.
Barry Leiba commented that it is very dangerous to measure
participation based on document authors; there are other ways to
measure participation. Ray Pelletier responded that attendance is
another way to measure participation.
Stanislav Shalunov brought up the issue of Friday afternoon meetings.
While it is declared that Friday is a normal day like any other, the
reality is that it isn't like any other. Working groups scheduled
for Friday have drawn the short straw. Is there a way to change the
calculation is to put something valuable on Friday. Russ Housley
responded that we have been trying to figure this out, and we still
don't have a great answer. Working group chairs continue to request
not to have Friday slots. Bob Hinden reminded folks that the plenary
used to be at the end of the week, and each AD would do a summary
of what happened during the week. He also reminded folks that the
IAOC doesn't make scheduling decisions and the issue should be
addressed to the IESG. A final, possibly not serious suggestion, was
to close the meeting with the social. Spencer Dawkins pointed out
that the technical plenary might be an enticement for people to stay.
Christian Vogt doesn't see a reason to keep folks for all five days if
the things they are interested in are not meeting. Bob Hinden responded
that having cross fertilization between working groups is good.
With no additional people standing in the mic line, Bob Hinden closed
the IAOC open mic session.
IESG Open Mic
=============
Russ Housley opened the IESG open mic session. All the current IESG
members came to the platform and introduced themselves.
Barry Leiba recalled an experiment with a single plenary, but he could
not recall the feedback or results. Russ Housley pointed out that
the result was the leadership was exhausted.
Pete Resnick brought up an ongoing issue related to the YAM working
group. This working group was formed explicitly to move established
email specification to Full Standard. The initial approach was that
documents would go with some explanation to the IESG first for an
initial reading (preliminary call on whether they are ready to go
forward or not a good idea). The problem is that we haven't figured
out a good process to do this. What can we do to make the process
better? There has been a lot of concern about the words commitment
or guarantee with respect to the initial reading. No one is asking
for hard guarantees, what we are looking for is some preliminary
review to make sure we are on the right track. Russ Housley responded
that the IESG has agreed to address this type of request as management
item on the IESG telechat. Pete Resnick asked for documentation other
than minutes of IESG telechat for current and future IESG members to
refer to. It should indicate that "these are the issues we reviewed",
"these are the things we thought were non-problematic", "these are
things that we thought, even though they were problematic, there was
enough legacy to go forward anyway", and so on. There needs to be some
record of the discussion so that when someone later brings up an issue,
we have documentation to refer back to that indicates what items had
been previously considered and the reasoning on those issues. Lisa
Dusseault asked about what happened with the idea of progressing
documents to full standard. Pete responded that they are wrestling
with the issues associated with it.
John Klensin commented that this whole notion started out as an idea
of a working group working with the IESG to progress documents.
At some point it turned into a wrestling match on how not to turn
things into full standard. We have somehow turned an idea about doing
something lightweight into something really heavyweight without
engaging in a dialogue. We need more engagement from the IESG to
progress. The idea here was to figure out how to engage enough to
get to full standard. Some group of people think that fooling around
with process could define an experiment. You folks should be
considering whether you want to engage on this or send a message to
the community that you don't want to pursue full standards.
Russ Housley responded that, in this case, the manner in which the
question was brought to the IESG surprised many of the IESG members.
Russ suggested dedicating one of the IESG Informal Telechats to a
process discussion on these points.
Loa Andersson mentioned a working group that has been around for 13
years with one RFC that is at Draft Standard. We don't have a process
problem, we have a mindset problem. There is too much work required
to get a document to Full Standard.
Ross Callon pointed out that in the Routing Area we have a whole lot
of Proposed Standards that the Internet runs on. It is extremely rare
to move something to Full Standard
Ralph Droms has chaired a working group for 20 years with two documents
at Draft Standard and the rest at Proposed. When considering progression
on the maturity ladder, Ralph was once asked by an AD "is that really
where you want to spend your time?" Jari Arkko indicated that sometimes
the motivation exists to do this when you have something else to do at
the same time.
Tony Hansen reminded the attendees that there was once a proposal
to move to a two step process versus the current three step process.
Also, we have a repository of interoperability reports. We do not
have a repository for deployment reports.
Tim Polk wanted to be clear that we didn't say we rejected the two
step process, we did come back and say we would prefer to do it this
other way. Hopefully we will work our way to something that works
for all.
Larry Masinter reminded everyone that several years ago there was
a working group called NEWTRK. He had an idea from that effort to
formalize the interoperability reporting in a way that would make
it less synchronous. You would document features and then document
interoperability. Russ Housley asked if there was an old draft in
NEWTRK that explains that idea. Lisa Dusseault pointed out that she
took ideas from Larry's earlier work to develop the recent BCP
published on interoperability report guidelines.
Russ Mundy pointed one example of being able to succeed to Full
Standard: SNMPv3. He is not sure where that motivation came from,
but we either ought to do the work the way we say we are going to
do it or change what we say about the process. Dan Romascanu replied
that the motivation came from the charter of the SNMP working group.
Dave Harrington replied that in his memory the motivation was 1) we
had to deal with SNMPv1 and 2) social engineering experiment to get
people to move to deploying SNMPv3.
Spencer Dawkins reminded folks that he comes to the IETF to do
technical work. In his opinion, process stuff gets in the way of doing
technical work. If you want people to believe the BCPs, they actually
need to say what we really do, don't make this hard, don't start a
General Area working group, don't redo NEWTRK all over again.
John Klensin spent some time reflecting on where we've gotten. RFC 2026
imposes almost no requirements on moving to Full Standard. The
requirement is wide deployment. We ought to be able to move these
things to Full Standard if protocol x is widely deployed and used. We
should be asking ourselves the question, are there incentives or
barriers to getting to Full Standard.
Lars Eggert pointed out that there are a new breed of TCP implementors
who are struggling with the existing documentation. There might be
clarifications needed. Just moving the document to Full Standard is
not enough in all cases.
Matt Mathis indicated that for most things it doesn't bother me at
all that these things haven't been pushed through. Eric Burger
commented that in reality there is no Full Standard category anymore.
Ed Jankiewicz pointed out that the Vatican named more saints this
year than the IETF named Full Standards. The Vatican doesn't make
saints, they recognize saints on their own merits. We are spending
too much time talking about the process and not the products.
Danny McPhearson suggested we impose a time limit on questions.
Larry Masinter asked do the documents accurately reflect what is
implemented and deployed? Lars Eggert responded why doesn't the
IESG get this question? What do you want the IESG to do? Larry
answered that he wanted a clearer process for defining a document for
Full Standard. Pasi Eronen reminded folks that the reason we are
getting these questions is because we badly screwed up the handling of
first draft from the YAM working group. We have a better idea going
forward.
Tony Li pointed out that being Full Standard has no practical
implication. It seems like work for no reason.
Matt Mathis suggested the Tao for the IETF needs to say something about
how the process actually works.
With no additional people standing in the mic line, Russ Housley
closed the IESG open mic session.
|
