IETF 80 Administrative Plenary Prague, Czech Republic Wednesday, 30 March 2011 Minutes by Mirjam Kuehne 1. Welcome Russ welcomes the IETF participants and presents the agenda. 2. Host Presentation Ondrej Filip, from CZNIC, represents the host organisation and tells the audience the third man in space was a Czech. (see slides). The IETF participants thank CZNIC for hosting IETF 80 and Russ presents a plaque to him. 3. IETF 25th Anniversary On 16 January 2011, the IETF turned 25 years old. (see slides) Russ introduces Mike Corrigan, Chair of IETF 1. He is present here at IETF 80 together with four others who were among the 21 attendees at the first IETF meetin. Mike shows the original notes and action points of the very first IETF meeting. Some of the action points are still ongoing today. (see slides) Jim Galvin: How many IETF meetings have the five attendees of IETF 1 attended in the meantime? None of them attended all of them. Mike St.Johns has attended the most; he has attended 76 IETF meetings. Russ shows a short video clip describing the work of the IETF and asking for support. ISOC helped put this video together, and it is now on the front page of the IETF web site. 4. Reporting IETF Chair Report by Russ Housley (see slides) Representatives from RIM, the host of IETF 81 give a short presentation to introduce Quebec City. Marc Blanchet set up a web site: http://ietf81.ca to share information about the meeting.    NOC Report by Morgan Sackett (see slides)    IAOC Chair Report by Bob Hinden (see slides) Bob Hinden makes two announcements: 1. We ran out of IPv4 addresses, and he can now stop making these graphs showing the numbers of IPv4 addresses going down. 2. IETF 84 will be in Vancouver! Google will be the host.    IAD Report by Ray Pelletier (see slides)    Trust Chair Report by Bob Hinden for Marshall Eubanks (see slides)    NomCom Chair Report by Tom Walsh (see slides) 5. Thank Outgoing IESG and IAOC Members All the outgoing IESG and IAOC members receive a plaque as thank you for their years of contribution. Many thanks to: Henk Uijterwaal -- IAOC: 2009-2011 Olaf Kolkman -- IAOC: 2007-2011 Alexey Melnikov -- APP AD: 2009-2011 Lars Eggert -- TSV AD: 2006-2011 Tim Polk -- SEC AD: 2007-2011 6. Introduce Incoming IAOC Members Dave Crocker and Bernard Aboba join the IAOC. 7. IAOC Open Mic Andrew Sullivan: was taking minutes for the IAOC and is now stepping down from that duty. He would like to encourage people to volunteer for this position. One learns a lot by doing this and you do a good service for the community at the same time. Especially when you are new to the IETF, this would be interesting. Sam Weiler: is disappointed how the IAOC handled the issues about badge checking in Beijing. We have not yet gotten a clear answer why this was done and who was responsible for it. Not asking for an answer now, just noting that this wasn't handled very transparently. Paul Vixie: Thanks Bob for the graph about the IPv4 address running out. Just wanted to remind people that the RIRs still have some IPv4 address in their pools. Anyone who has the power to increase the speed with which the network goes dual stack, please use that power. Eric Rescorla: +1 8. Introduce Incoming IESG Members Pete Resnick -- APP AD Wes Eddy -- TSV AD Stephen Farrell -- SEC AD 9. IESG Open Mic Andrew Sullivan: is interested in the decision to move the deadline for WG Chairs to request session slot. He noticed that some WG sessions run very long. Often very long presentations, some of which appear in more than one WG. WG chairs might be asking for even more time now with the deadline being moved. What are you going to do when everyone is asking for 2.5 hours? Russ: we talked about this. We will certainly look into the issue of repetitive presentations. It has been suggested that in case like this, the presentation is given in full only once, and then only a summary or abstract in other WG sessions. Russ: People were asking to get the agenda out earlier. Therefore the WG deadline for session slots was moved up. Andrew Sullivan: clarifies that he really appreciates getting the agenda out early, is just terrified that every slot lasts 2.5 hours. Robert Sparks: This can be solved bottom-up: He suggests individuals in the WGs push back when they think a WG is not run efficiently. Lars Eggert: ADs now get copied on the time slot requests and can step in if they think it is too much. Thomas Narten: In some WGs there is a lot of activity but it doesn't move the WG forward. Its a broader issue, needs a culture change. It needs a combination of top-down leadership and bottom-up involvement. Pete Resnick: Do you want us to start coming down on people doing presentations? Because I'm all for it. Thomas Narten: Don't just pick on powerpoint. This is only a symptom. There is a whole range of things that could be done to improve that situation. Adrian Farrel: in the open routing open area meeting it was discussed that something needs to be done about this. It was suggested that authors don't present everything that is in their draft already. Slide ware was suggested. Maybe something between an abstract and an introduction in PDF is needed. Dave Harrington: We have started to request the chairs to focus the first part of their meetings on charter deliverables. So we can try to make sure that we're making progress where we're where supposed to be making progress and that the additional slideware that is there for presenting new ideas tends to be pushed to the end. So if we run out of time, it's the new stuff that gets cut, not the deliverables. Keith Drage: This new deadline basically provides one extra week of conflict resolution. Maybe we don't need that if we find another solution to solve scheduling conflicts between WG slots. Dave Harrington: We are considering to review the primary and secondary conflict lists. Russ: When you submit a request for a WG slot, please review all the information presented in the form and make sure it is still current. It defaults to the information provided with your last request. Lars Eggert: We started to have a conference with the entire IESG to go over the initial agenda the Secretariat provides and try to solve as many conflicts as possible. These calls take a lot of time, but we seem to have fewer conflicts. Dave Harrington: And BoFs are now scheduled separately from the Area meetings. Leslie Daigle: was in the routing open area meeting today and appreciated all the comments made there. Wants to second Thomas Narten's point that powerpoint is only a symptom. There are other issues where WGs could be run more efficiently. WGs needs proper meeting planning and proper meeting running. Dean Willis: managed to go to a few lunch-time presentations and meetings and really enjoyed them. Good to learn new things. Can we formalise this and have dedicated time slots for tutorials and introductory sessions? Pete Resnick: We do have open area meetings. When would you like to do something more complex? Jari Arkko: On Saturday we offered a useful tutorial on the Internet of Things. Fred Baker: is concerned that the people who sit in these sessions assume they are being trained. Hopes that the audience is providing a different perspective than the speaker. Lars Eggert: agrees with what Jari said, it was a very good tutorial. Russ: we do some tutorials on Sunday. Based on this discussion, there seems to be a call for more. Is that what I am hearing? Spencer: really liked the Saturday tutorial on the Internet of Things. On the topic of effectiveness, if you're not presenting your draft at the IETF, that's actually a win. If you could be having conversations where people discuss your draft, they understand your draft, they accept ideas in your draft. Your draft might even be picked up as a working group item, without having to wait on the average two months to get to an IETF meeting. Andrew Sullivan reading a jabber comment from John Klensin: +1 to Andrew. The earlier requests come in, the earlier people can estimate how much time they will need accurately. And the more requests we get for long slots is a precaution. It is sometimes just not possible to make a good estimate a month in advance. An accurate time estimate requires knowing what drafts will be posted and what issues will be resolved for the first time before the meeting. Tim Polk: those are things that a WG chair should actually know a month in advance. Peter Saint-Andre: Would even go further and say that is something an AD should know. Eliott Lear: Maybe we can charge by the number of words aired at the microphone. Was blown away by the quality of the presentations in the TSV open area meeting. Suggests people to listen to some of the presentations in this group, two in particular. Wes Hardaker: Attended the nfs4 working group this week. Sat in the back of the very crowded room and learned absolutely nothing. Don't expect to be educated. We need more cross-area work. That means we will need more tutorials. Idea: we have a two-tier model. Second tier should have a requirement that you hold a tutorial. Curtis Villamizar: a change that was made ages ago was that a deadline was put on draft submissions. This is still very useful. This deadline, especially for individual draft-00 submissions should be moved back, so that people have more time to actually read the drafts. We may want to consider a deadline for slides to get out ahead of the meeting, then people could actually look at them prior to the meeting. Russ: This was suggested numerous times. But there is a lot of push back, because people want to include their last minute ideas. Curtis: maybe on a per-area basis? ??: This is the WG chairs responsibility Lou Berger: really likes both of Curtis' points. It should also not been allowed to do a -00 draft before the first cut off date and then a -01 draft before the second cut-off date. Also fully supports the notion of moving those individual submissions earlier to give people time to look at them. Tobias Gondrom: as a WG chair, he is very conscious that people submit as early as possible to allow many people to read the document. However, he would not suggest to extend the deadlines. Keith Drage: Stick to the deadlines as they are. Discussion should take place before the WG meeting on the mailing list and not during the physical meeting. Adrian Farrel: This really is the job of the WG chairs. Because the education will come through hard knocks, and next time around people will get their drafts out and get them discussed. Lou: following up to his earlier comment, he finds it useful to get slides out earlier. If those slides include some backup material that is not meant to be shown, but people can look at it for educational purposes, that would be helpful. Dave Crocker: We have a tendency to focus on symptoms. The problem is that we are supposed to do our work on mailing lists, not in the physical meeting. Getting drafts in at the last minute, means the WG hasn't been doing what it was meant to do. Thomas Narten: Likes the discussion. Having an earlier submission for individual submissions would make sense indeed. That said, I am not sure he is in favour to make a strict rule about this. What we need fundamentally is WG Leadership! Top-down leadership from the ADs would be helpful. Without that kind of backup some WG chairs might be hesitant to implement stricter and more efficient rules in the WG sessions. It might be useful to set up a team. This is not rocket science. There are WGs that do an excellent job. Others do a less good job. These cases we need to fix. Adrian Farrel: Just to clarify: When I criticised WGs for not doing their job, I am fully aware that I am also to blame. Andrew Sullivan reading a jabber comment from John Klensin: Has the IESG considered letting WG chairs create rules about regarding posting, for example, if a working group wants to impose an early limit or WG chair, for a second posting within the one month time limit, why not let them? Russ: Well, they certainly can do it, by denying a slot to somebody who posted past the internal deadline. Paul Vixie: speaking against the idea to get slides in early. Slides should be teasers one meant to speak to. If they are understandable in itself, they are probably too long. Bob Hinden: Current deadlines were set because I-D processing was manual. Now that it is automated, we could probably move the deadline closer to the meeting and get more work done. Russ: that was not the reason that I was told. I was told that the deadline ensures that the newest material was out soon enough before the meeting for people to read it. Bob: not sure if it will be useful to move the deadline back more. Russ: new I-D submission tool fixes some issues. The hope is that there will be no manual work involved anymore. Pete Resnick: That said, at these earlier times, there were not as many drafts. Jari: Some of the drafts we are processing, only need a small fix to get the draft improved. Russ: Any AD can get a document posted during the black out period. Gonzalo Camarillo: We are very well aware of the problem, but individual WGs need to be managed in different ways. Curtis Villamizar: disagrees with Bob Hinden. Remembers times where we had to read tons of I-Ds just before the meeting. Scott Bradner: RFC 2418 says: All relevant documents should be made available two weeks before the meeting. So, yes, Bob is right, the reason was that we had to process them manually. Going to more than two weeks is possible according to the RFC. Going to less than two weeks, we would need to change the rules. John Parello (?): there are groups that hold interim meetings. Even in the case of individual drafts, it is difficult to be categorical about it. Suggests to keep things as they are. Keith Drage: The dispatch WG for instance has deadlines well in advance of the IETF deadlines.