Skip to main content

Minutes for IESG at IETF-91

Meeting Minutes Internet Engineering Steering Group (iesg) IETF
Date and time 2014-11-13 03:10
Title Minutes for IESG at IETF-91
State Active
Other versions plain text
Last updated 2016-02-16

IETF 91 Administrative and Management Plenary
Minutes: Amy Vezza, IETF Secretariat

1. Welcome - Jari Arkko

2. Host presentation - Dave Ward

3. Reporting

      - IETF chair - Jari Arkko

Statement on the VISA issues for IETF 91:

      - IAOC chair and IAD - Chris Griffiths and Ray Pelletier

      - IETF Trust chair - Tobias Gondrom

      - NomCom chair - Michael Richardson

4. IAOC open mic

John Levine asked about the fee increase for meetings. He stated when looking
at the IAOC report, the slides indicated a projected flat revenue for the next
year. If that is the case, why the fee increase?  Ray Pelletier mentioned that
the revenue only looks flat because the projected attendance at two of the 2014
meetings has been higher than expected, and at one of the meetings it is lower
than expected. The differences cancel each other out in that way. He went on to
talk about meeting venues that require meeting space fees, one meeting in 2014
(London) and two meetings in 2015 (Prague and Yokohama), so projected expenses
will be higher for 2015. In 2016 all three meeting venues will require meeting
space fees, so expenses will be even higher.

Lee Howard talked about hotel arrangements for this meeting, and that the IAOC
took a lot of flack for the choice of venue and place (Honolulu, Hawaii). He
mentioned he has no issue with the venue, that he likes it very much.

Bert Wijnen raised the point that students don't have the funds and have the
opportunity to enjoy the IETF Social Event. He mentioned he'd sent email asking
about the possibility of everyone paying slightly more per ticket to allow the
students to attend for free. Ray Pelletier mentioned that the fee for the
social goes back to the host to help defray their costs for the event. He
mentioned that perhaps future hosts can offer a different pricing tier for
students to attend the social event, but ultimately it is up to the host.

Lars Eggert followed up the question from Bert with the comment that at SIGCOM
students pay less for the same experience as everyone gets there. He mentioned
he would count the Social Event as part of the experience, so it should be
discounted if possible.

Barbara Stark asked, supposing students got the discounted tickets, if it would
become policy that students could not resell their tickets at full price.

5. IESG open mic

Wes Hardaker mentioned that as a document author, getting really good feedback
that might change the scope of a document from the IESG at the last moment
should trigger a re-review in the working group but rarely does because of the
exhaustion from all involved. He mentioned that getting that kind of in depth
review would be valuable earlier in the process but is at a loss when such a
review could happen and who would have the time to do it. He would like to know
the perspective of the IESG. Jari Arkko mentioned that the IESG has observed
this, and the policy is that the review should go through the working group. 
Wes mentioned he never received any kind of instruction from an area director
that a change in text is so significant it should go back to the working group.

Barry Leiba mentioned that the IESG has sent documents back to the working
group. He mentioned that the commenting AD, the responsible AD, the document
shepherd, and the working group chairs should all be aware of what changes are
being made to a document and if those changes are significant enough the
document needs to go back to the working group. He mentioned that they have
missed a few that should have gone back, but there is a current document that
should be in Last Call soon on extended document shepherding that gives
guidelines to document shepherds to keep an eye on this issue at all stages of
the process.

Pete Resnick mentioned a subtle tool change in the datatracker that includes
the working group mailing list on the review email and that should help keep
the working group involved in the discussion.

Spencer Dawkins mentioned the IESG is also including working groups in the
working group charter changes, and the review discussions of new and revised
working group charters.

Jari Arkko added that all of this transparency makes it everyone's
responsibility to make sure the changes that are suggested by reviewers are
vetted by the appropriate parties.

Andrew McGregor mentioned he has both taken a review back to the working group,
and been instructed to take a review back to the working group. He added that
in both cases the second WG Last Call resulted in silence and the WG was
perfectly happy with the change.

Randy Bush mentioned he was worried about items going back to working groups
and reopening contentious discussion.

Wes Hardaker mentioned he has seen documents go back to working groups, he
didn't mean to imply it never happened. For major changes, he's seen the
document be sent back, but for some more minor changes there was more of a
"just do it" attitude that was concerning. He's seen some substantial technical
changes that should have gone back and did not and he'd just like the community

Jari Arkko added that the area directors review a significant number of
documents in a telechat cycle and sometimes don't remember to take those
reviews back to the group that produced the document.

Barry Leiba noted that an AD putting a discuss position on the document means
the AD wants discussion. Sometimes the response is a change in the document
with no discussion, and he'd prefer the discussion take place. With the change
in the notifications he hopes to see more discussion take place with the
working group.

Spencer Dawkins mentioned that the IESG is also working on the types of
discusses that would require the working group to review the document again.

Michael Richardson brought up the subject of the posting of working group
agendas and the discussion on the mailing list on the agenda deadline compared
to the draft deadline. He suggests moving the agenda deadline earlier than the
draft deadline. His suggestion is for the convenience of the draft authors
knowing their documents would be discussed or not. He mentioned that if WG
chairs didn't know at least a month before the meeting what the agenda for the
meeting was going to be they should have their meeting slot cancelled or
significantly reduced.

Ted Lemon agreed that he's seen working groups ask for a two and a half hour
session and only use one hour.  He'd like to see agendas from working groups
before the allocation of time/space on an agenda.

Alia Atlas mentioned that the time management and project management is really
important for WG chairs to have a firm grasp of for their group. She went on to
say that face time is really valuable and it needs to be spent of document s
that have active interest in the group. And it does mean more management, and
possibly more virtual interim meetings and more clearly defines milestones to
be able to work on items between IETFs so work gets accomplished. She likes the
idea of having the WG Meeting agendas due the week before the draft deadline.

Benoit Claise mentions he holds a pre-IETF teleconference with his chairs and
they look at the agenda as a whole. It is something the OPS area has done for

David Black talked about the draft deadline and how many times a draft 00
document shows up right at the document deadline and is accompanied by a
request for agenda time. If the agenda is required to be available in advance
of the draft deadline it would make the initial document (version-00) also show
up well in advance of the deadline. He thinks this would give authors and
chairs more time to discuss changes needed to documents so they can get a
review spot in the working group meeting.

Brian Haberman talked about a practice he's used and seen used in some working
groups where the chairs set out the agenda in terms of priority. First priority
goes to working group documents, second priority to individual documents that
have seen discussion on the WG mailing list, and then brand new documents as
time allows. This allows chairs to prioritize work that needs to be done.

Fred Baker mentioned that the change of moving the date for the drafts closer
to the meeting makes it harder for WG chairs to set meeting agendas. He talked
about requesting agenda items two months ahead of the meeting and still ended
up getting ore than one 00 drafts at the deadline. He discussed setting a
deadline in advance of the draft deadline just to keep the agenda manageable,
but as a chair he'd prefer the deadline be set by the IESG.

Barry Leiba mentioned the IESG would prefer that be set by the WG chairs
because one date doesn't work for every group or every situation. He said
they'd rather the working group chairs set the date that works for the group
and their management mechanisms.

Brian Carpenter brought up a document that talks about if operators feel their
input is welcomed in the IETF and if the statistics quoted in the draft are
true they are disastrous. He wants the IESG to read the document carefully and
take action if necessary.

Alia Atlas mentioned the document is a good one, and she has discussed the
conclusions with the authors and started brainstorming ways to bring operators
in to the IETF and how to make sure they know the IETF does welcome their
input. She recommends anyone interested talk to Chris Grundemann who is
gathering ideas on how to make it work better in some of the operator fora.

Joel Jaeggli talked about outreach being important in both directions, but he
doesn't believe all operators will ever agree.

Alissa Cooper agreed with Joel in if the IETF acted on all input from everyone,
the IETF would not be as useful.

Benoit Claise also read the document and had a conversation with Chris on ways
to raise visibility. He has also discussed tool improvement with Henrik for
better document tracking. He mentioned that at the YANG session he was happy to
see a newcomer with a document on YANG models.

Jari Arkko mentioned the importance of remembering the networks that are in
use, rather than just the theoretical network.

Leslie Daigle thanks Jari for bringing up the IANAPLAN WG in the slides
presented earlier and wanted to share information about the document discussed.
The document has completed WGLC and with some work on the part of the authors
they expect it to move to Publication Requested in the next few weeks.  She
requests anyone who was part of the discussion and had strong opinions about
the way forward to share thoughts and comments on the mailing list.

Jari Arkko was interested in how many people present were following the
document in question. A show of hands had a "good chunk" of the room interested.

Philip Hallam-Baker commented about the issue brought up about -00 drafts and
hopes -00 drafts will not be considered the problem rather than part of the
solution. He mentions that often when he raises issues on a mailing list WG
chairs will tell him to "write it down in a draft". If there is an inflexible
set of deadlines as in the past when -00 documents had a deadline a week before
the revisions and when authors would submit empty documents as placeholders for
revisions-to-come. He urges the IESG not to create inflexible rules.

Pete Resnick noteed the current rules can be circumvented with AD approval.

Adrian Farrel commented that what was not needed was new rules, but better WG

Kathleen Moriarty added that there are more documents submitted immediately
before and after an IETF meeting, it would be nice if there were more documents
submitted between meetings, having the submission of drafts distributed a bit
would gain them better reviews.

Benoit Claise mentioned how useful it is to have WGs have more virtual interim
meetings. It has helped the work advance in the OPS area, and he thinks other
groups would benefit.

Jari Arkko agreed that is an important part of the work.

Bob Hinden mentioned the issue of drafts and agendas is a management problem of
ADs managing WGs and the IETF doesn't want the ADs to be reviewing and
approving agendas. WG chairs should be doing that and if they are not doing it
effectively then that should be fixed, either by getting new chairs or training
them how to do it.

Bob Briscoe mentioned on the topic of -00 and agendas. He thinks if the
deadline for 00s before the agenda deadline he thinks there will be more
conflated requests for agenda spots as people may think they can get more done
before the meeting than they are able. Which means they will need to be
stricter about taking items off the agenda that don't make the document
deadline. Which means less useful agenda information earlier.

Jari Arkko agreed but mentioned that agendas always change over time and get
more refined the closer to the meeting.

Bob noted perhaps a better question for chairs to ask is what the probability
is of the authors needing a time slot, not do they need a timeslot.

Kathleen Moriarty noted that the WG Chair can be sure there is a document
before granting an agenda timeslot and there is agenda bashing available right
up until the meeting starts.

Ted Lemon mentioned that the topic of a document must be germane to the working
group. That the topic can be discussed even before there is a document.  If
there is a document that doesn't need to be discussed, it does not need an
agenda timeslot. Chairs need to determine what is trying to be accomplished
rather than which documents will be discussed.

Alia Atlas stated that process is a tool to help people and it shouldn't be a
strict mandate. She mentioned that the chairs and ADs cannot do it alone, and
everyone needs to help in that regard. She went on to list interim meetings,
design teams, and focused discussion on the mailing lists as ways to drive the
change in perception.

Pete Resnick added that this discussion topic has been discussed at plenaries
and on mailing lists before.  He mentioned that wholesale change is not a good
idea, but that chairs should discuss issues with other chairs in their areas or
with their ADs to develop strategies that work with their particular working

Joel Jaeggli mentioned a "lousy idea" that if the issue is structured around a
problem of incentives, WG Chairs can construct a futures market for
presentation slots and they will either get more expensive or cheaper as the
deadline for agendas gets nearer. <laughter>

Dan Romascanu did not like the discussion surrounding the deprioritizing of new
ideas and especially -00 drafts from new participants. He proposes an
experiment, to provide ad-hoc space and time for new document authors to be
able to discuss their -00 document to those interested in the topic. It will
help the IETF encourage new ideas and help new participants find working groups
who may be interested in adopting their document as a work item.

Jari Arkko mentioned this a good point. He mentioned a lot of this type of
discussion happens in hallways and other areas rather than in WG meetings, but
maybe an experiment like this would be useful in the future.

Ted Lemon mentioned that in response to the WG shopping, sometimes new
documents don't get shown to the right set of working groups, so it might be
good to do that sort of thing in the area open meetings to hear about new
documents and discuss where they might fit.

Brian Haberman talked about a few working groups that have adopted the NANOG
model of having lightning talks. The last few minutes of their meetings are
given specifically for new -00 document authors to give a short summary of the
document to the working group and invite discussion after the meeting.

Dan replied he was looking for something more in depth that might have a wider
scope than a few minutes at the end of a WG meeting.

Brian mentioned that the model Dan first discussed reminded him of poster
sessions at academic conferences. To give people the venue to discuss their
ideas on a high level. But the IETF participants must be willing to attend such
a session.  He went on to say that giving people an opportunity to open the
dialogue is important.

Spencer Dawkins added that having a better venue to assimilate new people and
new ideas rather than having them present in working group meetings might have
merit and be a better use of time for the working group and for the document
author looking for a home for their new draft.

Jari Arkko mentioned this seems to be a good idea and should be thought about

Iljitsch van Beijnum mentioned that sitting in working group meetings and
listening to the people in the meeting debate small points about a document was
not the best use of time. He added that the discussion of the new ideas, and
new documents is much more interesting to him.  He thinks finding a balance
between the two would be useful in the long run.

Wes Hardaker talked about a document he wrote he couldn't get people to review
until he was granted a few minutes at the end of a working group session that
sparked interest in the topic. He thinks the IETF needs to find balance. He
mentioned that face to face time should be reserved for discussions that have
not been solved via email, questions to help resolve issues, and new work.

Alia Atlas mentioned that one on one conversations in the hallways was an
immeasurable asset for finding people interested in a new document. She added
that new documents had to pass "sanity checks," and the ADs relied heavily upon
the WG chairs to filter out the bad ideas.

Lee Howard pointed out that as a community we are still allowed to discuss
documents on the mailing lists, as the discussion made it sound like that
didn't happen.

Richard Barnes mentioned DISPATCH as a RAI area working group that discusses
new ideas in the RAI area.

Randy Bush mentioned in relation to the operators discussion earlier that
things have changed and they are better and there were a significant number of
operators here where ten years ago they weren't.

Chris Grundemann mentioned that the document about operators was written as an
exercise in how to get more operators involved in the IETF. It became a
blueprint for how to make access and participation easier for them, but also
may help get others, not just operators involved.

Other resources:

NOC Report:

IANA Report:

IETF Secretariat Report:

RFC Editor Report: