Administrative Plenary Wednesday, 10 November 2010 Minutes by Alexa Morris AGENDA 1. Welcome 2. Host Presentation 3. Postel Award 4. Reporting - IETF Chair Report - NOC Report - IAOC Chair & IAD Reports - Trust Chair Report - NomCom Chair Report 5. IAOC Open Mic 6. IESG Open Mic ADMINISTRATIVE PLENARY 1. Welcome by Russ. (http://www.ietf.org/proceedings/79/slides/plenaryw-0.ppt) 2. Host Presentation (http://www.ietf.org/proceedings/79/slides/plenaryw-8.ppt) Warm welcome to all IETF participants, it is a great honor to host IETF 79 in Beijing, the first such meeting in China. And thanks to Tsinghua University's two co-hosts: Internet Society of China and CNNINC. The three co-hosts each receive a plaque from the IETF Chair. 3.Itjun Award (http://www.ietf.org/proceedings/79/slides/plenaryw-1.ppt) Itojun was an extraordinary person who passed away three years ago: he was a member of the IETF community who was dedicated to IPv6. This years award was presented to Bjoern A. Zeeb, from FreeBSD for his work in making IPv6 a first class citzen in open source UNIX world. Bjoern thanked the community for the award. This is his first IETF meeting, but he hopes to be back. He has been on implementation side and a consumer of RFCs, and now, understanding what it takes to get an RFC published, he really appreciates the hard work the IETF is doing. 4. Reporting Russ Housley: IETF Chair Report (http://www.ietf.org/proceedings/79/slides/plenaryw-2.pdf) Anna Filinova: IETF 80 in Prague (http://www.ietf.org/proceedings/79/slides/plenaryw-4.pdf) Xiaoliang Zhao: NOC Report (http://www.ietf.org/proceedings/79/slides/plenaryw-7.pdf) Bob Hinden: IAOC Chair Report Ray Pelletier: IAD Report (http://www.ietf.org/proceedings/79/slides/plenaryw-3.pdf) Marshall Eubanks: Trust Chair Report (http://www.ietf.org/proceedings/79/slides/plenaryw-9.pdf) Thomas Walsh: NomCom Chair Report (http://www.ietf.org/proceedings/79/slides/plenaryw-5.pdf) 5. IAOC Open Mic (with the IAOC on stage) Introduction of IAOC: Bob Hinden, Henk Uijterwaal, Marshall Eubanks, Russ Housley, Lynn St. Amour, Ole Jacobsen, and Eric Burger Unknown: This venue is fantastic, and some of us were talking during the reception about Hawaii as a meeting location. Bob Hinden: Yes, it is definitely one of the places we have in mind. It came up at the IESG breakfast on Monday and they were very much in support of it, so it is certainly possible. Mary Barnes: I like the results of the survey about what makes a good venue. Also, I have had some wonderful meals here and in particular I want to thank Chef Eric. Sam Weiller: I've noticed that people are turning away people who don't have badges. This is a notable change in practice and it was announced with no explanation, no advance warning and with no request for community input. Bob Hinden: There is a requirement in an RFC, I don't remember which one, to wear badges. Badges have been checked at previous meetings occasionally, usually at the Terminal Room. The local hosts at this meeting required that only registered attendees be permitted in the meeting room areas. Sam Weiler: Whether or not we've checked when it comes to the Terminal Room is irrelevant when it comes to meeting rooms. And I'm still wondering why you didn't tell us in advance. Springing on us is not good. Bob Hinden: I think if we understood it would be an issue, we would have talked about it earlier. Kurtis Linquist: I noticed recently that the agenda on the ietf.org site doesn't always contain useful information and the tools page agenda contains more useful information. How are the newcomers supposed to find it the tools version of the agenda? Russ Housley: We know about this and are working to figure out the best solution. We are trying to figure out if a couple of different views of the same data would be useful. On Saturday the Code Sprint worked to provide a single button where you can download the agenda and the presentations for each working group session, with one click. Kurtis Linquist: I understand that people follow different WGs, and it would be good for the agenda to contain a link so that people could read the relevant drafts. Wes George: For this meeting, the hotel information was not posted at the same time as meeting registration opened, which is a change. And less than a week after hotel information was posted, all of the lowest rate rooms were taken. And when I pointed that out to organizers, mysteriously some of cheaper rooms became avalable. Is there a formal guideline about how many rooms of a certain price are available? And also, are you indexing it to something like the federal per diem in the U.S.? What is the expected maximum hotel room price range? Ray Pelletier: When we negotiate a contract, we try to get a certain amount of rooms on peak night, and we try not to go above 500. There usually isn't a difference between when the hotel and the meeting registration announcement, and I'm surprised that there may have been. It is true that we sometimes hear from members that rooms were not available, and we contact the hotel and they open up more rooms. Wes George: Where are you looking to provide a guideline on room rate pricing? Ray: We try to negotiate a room rate below market, and one that usually includes internet access. We also try include in the contract that the hotel can't offer a rate that's better than what they give us during the same week. But there is no formal indexing against a government per diem. Wes George: I might recommend that you consider that, because I've spoken with several folks who normally attend and they aren't here this time because it was too expensive. Barry Lieba: Since EKR isn't here to comment, I will say that there is now a new standard: we need dim sum at all afternoon breaks. Bob Hinden: I enjoy it when the food is local to wherever the meeting is taking place. Pete Resnik: Echo Barry's comment -- the hotel did a really good job this week. You said during your talk that you went to the IESG about the 1-1-1 and they approved it. Was this getting their buy-in as important community members or is the IAOC falling into a pattern of thinking that the IESG needs to approve things they do? Bob Hinden: Good question. It was suggested to me that the meeting rotation policy is actually an IESG issue. Pete: As a general rule, the IAOC should be doing the administrative work that eningeeers should not be doing. Unknown: Have you already signed a contract with the hotel for next November meeting? There were conflicts with another meeting, can we reconsider the date? Bob Hinden: We publish the dates for the meeting well in advance to avoid conflicts and we can't help it if another group decides to hold a meeting over the same dates. Ray can speak to the state of the contracts. Ray Pelletier: We have contracts with Taiwan International Convention Center (TICC) and with several hotels in the immediate area. Unknown: This is a comment on the tools style agenda and I agree that if we can get everything in one click it would be better. I would also love to have a basic IETF preparation checklist, to make sure that I don't miss anything before I leave to attend a meeting. Everyone will have their own minor differences, but a general checklist might be helpful. I will volunteer to create this. Aaron Faulk: +1 about hotel comments. And I'd like to propose an experiment for the next IETF: I would like to have beer available for purchase during the breaks in advance of the plenary. Stuart Cheshire: I occasionally find there's a mailing list I meant to subscribe to but forgot to do so, and so missed messages. It would be nice if there was some way to have it resend recent messages. Tim Shepard: You can use FTP. There's the dot mail file. 7. IESG Open Mic (IESG on Stage) Introduction of IESG: Ralph Droms (INT), Jari Arko (INT), David Harrington (TSV), Stewart Bryant (RTG), Robert Sparks (RAI), Russ Housley (GEN), Sean Turner (SEC), Tim Polk (SEC), Adrian Farrel (RTG), Alexey Melnikov (APP), Peter St. Andre (APP), Gonzalo Camarillo (RAI), and Dan Romascanu (OPS). Ron Bonica (OPS) was absent. Leslie Daigle: I'm glad that you are planning updates to how the agendas are set, and deadlines, because it's very important for people who are coming to be able to plan. However, the deadlines are not always followed and this week there were a lot of working groups that didn't have agendas posted in a timely fashion. We have to take the deadlines seriously. Russ Housley: I agree. We have asked WG Chairs to consider their agenda items when requesting a meeting slot. I would also encourage WG Chairs to make sure that the agenda is published well in advance and that the presentations are available before their working group session starts. Gonzalo Camarillo: We started experimenting with this last time in RAI. We asked chairs to prepare an agenda before the scheduling happens. We got some push back, but we really want chairs to go through this thought process and get to a rough estimation. We don't want everyone to just ask for the same time amount of time they requested last time. David Harrington: In the WG chairs lunch today, we got feedback saying that we were asking for them to plan too early. But I think that WGs go through life cycles and most working group chairs should know what their meeting is going to be about 3-4 months from now. Spencer Dawkins: Is there any guidance you would like to give us about BOF requests? Jari Arko: When you come to us with a BoF request, please come early. Discuss it widely, and discuss it with the ADs as early as possible. David Harrington: When you come to us with a BoF proposal, we look for a clearly defined scope and some community interest. We want to know that there are people who are going to do the work. Spencer: As an IAB member who also gets to watch these things, I love having you guys providing more guidance to the community. My experience in the last couple of IETF cycles is that the least well formed BoF requests take up the most IESG time. Dave Crocker: I thought that David Harrington's comments touched on a larger issue: we have people managing groups who are not managers. WG chair training is a good start, but we may need some advance training. Many of the people in this room are good at training; the ones who have a track record with difficult working groups are probably the best teachers. Spencer Dawkins: Ten years ago we typically opened the WG chairs training up to anyone in the community who thought that they wanted to know more about it. If that's still true, then I would encourage people to participate in the WG Chair training before they are WG chairs. Gonzalo Camarillo: In the RAI area we discovered that we didn't have a consistent definition of milestones, which made it very difficult. So, this is just to say that at a certain point it would be good to go back to the WG chairs training. Harald Alvastrand: I appreciate the thought that goes into milestones and deadlines but I'm also worried. I started working on an effort in Maastricht and then continued the discussion in October. After that October meeting, I found that the BoF deadline had passed and also the deadline for the dispatch working group had passed. We are looking forward to Prague but I'm concerned that we are so hung up in deadlines and procedures that we are losing flexibility. David Harrington: I mostly agree and I certainly believe that we need to be flexible. If there is a little bit of flexibility in the schedule, a WG chair can always go to their area director and let them know that the topic has proved to be hotter than expected, and make a request for time. I believe that's what happened this week. Gonzalo: I get approximately 20 requests per IETF from people saying that they missed the deadline, but that they would really like their presentation to be heard. And I cannot do that. But in principle I agree that we need to be flexible. Jari: its also important that we consider priority between working groups and within the working group; we just have to prioritize. Pete Resnik: I want to talk about BOF scheduling. BOFs are the principal kinds of groups that you want cross area review from and you want as many people to participate in that cross area review as possible. But putting all of the BOFs one on top of one another guarantees limited cross area review. I'm trying to understand what motivated this proposal and since there are a lot of people in the room, I'd like for others to give input. Russ Housley: I'm putting up the slides that address this. Do any ADs want to talk about why this was discussed in the IESG? David Harrington: We would love to have BoFs be attended by people from multiple areas, although there are always going to be scheduling problems. The specific BOF you mentioned touched on a bunch of different things, and did not have a clear scope. We told the organizers that they needed to give us a better scope, that it did not currently have a scope that could be reasonably assigned to one area. We talked about a number of different approaches. We can pick one time of day and schedule all BOFs at that time. But if you happen to be in a WG session, you can't attend. We considered having one time slot that held all BOFs, which would increase conflict between BOFs but reduce conflicts with other sessions. It would also force the IESG to prioritize the BOFs, because we only have eight viable slots at one time. And that is what we ultimately proposed. We can try to do them diagonally, if really want to avoid conflicting with area meetings. Pete Resnik: Conflicting with WGs is going to happen. But the area issue conflict is a big one and it doesn't seem problematic to me to schedule all area meetings on top of one another. Area meetings are specific but BOFs are nebulous, and you don't know specifically what area that topic will fall under. So I think it's much more important to avoid conflicting with area meetings than WG meetings. I want to know what the trends are in other areas and I cannot attend all their WG meetings so the area meeting is critical to me. Jari: There are tradeoffs. Are we going to have conflicts between BOFs or between BOFs and WGs? It's a tradeoff between inconveniencing those people who don't want to be here on Friday or Monday, versus the ability to do scheduling at a very late stage. So if you want the flexibility to schedule some meetings at a very late stage, you'll also have to manage the fact that some people might miss that meeting because they already have flight arrangements. Tim Polk: One reason we made this proposal is because we'd like to see BOFs earlier in the week, so that people would have the remainder of the meeting week to talk about the ideas presented in the BOFs. Stewart Bryant: We are currently at the limit of being able to create these IETF meeting schedules -- they are very complex. Some people want to focus on the newer ideas; some people want to focus on completing the work already in production. This particular approach prioritized existing work over new work, the feedback we are getting gives us more data to work with. Tony Hansen: At the WG Chairs meeting we talked about how to use face-to-face time more effectively. Thinking back, the most effective sessions I've attended were those where people discussed issues with each other, rather than just made presentations. This does not work in meetings of 400 people, but it does work in meetings of 20-30 people. Perhaps we could experiment with having one small meeting room configured in a roundtable format, to support discussion. We could have a couple of mics to share. There could still be a screen, for presentations, but the setup would promote face-to-face discussion. People could make a special request for that room, via a checkbox in the session request tool. Russ: Perhaps we can set up a design team room in Prague, with a flipchart and/or whiteboard. Mark Wood: As you are trying to encourage people to attend BOFs, perhaps you might draw the diagonal the other way. Russ: The diagonal was fairly random and the other way might be useful as well. David Harrington: Is it possible to get whiteboards in several other meeting rooms as well? Russ: We can look into it. Adrian Farrel: I don't think chairs are screwed to the floors. So if you want to meet in the round, move the chairs. Hannes Tschofenig: The challenge with using a whiteboard or flipchart is that it impacts the ability of remote attendees to effectively participate. I would suggest that there might be a better approach that would enable remote attendees to more fully participate. Dan Romascanu: I believe we should create room for design teams. And electronic versions of whiteboards and other collaboration tools are key also so that we allow remote participation, because remote participation is a fact of our life. Andrew Sullivan: Are there criteria for delegations inside the ietf.org zone and are there criteria for removing delgations from the ietf.org zone? Russ: There are some domains that are subordinate to ietf.org, such as ops.ietf.org and tools.ietf.org. However there has never been a constant criteria over the life of ietf.org, nor has there been discussion of how subordinate domains would be revoked. I will put this as an agenda item for a future IESG telechat. Dave Crocker: Tony Hansen's idea surprised me, because he and I and many others participate in WGs that operate much as he is suggesting. The thing about WGs and BOFs is that with WGs, they very quickly have a population and that population can assert what conflicts it wants to avoid. BOFs do not have a population yet. I'm thinking that we could add a "first call" mechanism, which would involve circulating an announcement about BOFs, asking for people to indicate which conflicts the BOF should avoid. This will provide more data in the scheduling. Russ: I disagree with you. As BOFs are currently structured, we pick a BOF chair and they are informed about who wants to make presentations, who has ideas. So, we are not entirely naive about the population. Dave: That is a biased sample. Russ: I understand your point. It is not as though we are totally unaware. Tobias Gondrom: I agree with Harald's point about needing flexibility in scheduling if something legitimately comes up late. What if we have a flexible slot on Friday afternoons, so that if it genuinely comes up late perhaps you don't mind taking the Friday slot. Russ: We had a similar thought, and this is a topic for a future IESG telechat. We would like to provide an opportunity for lightening talks. Mike St. Johns: This goes back to your earlier presentation, about statistics. One, we have a net increase of two working groups. Part of the problem with respect to BOFs and WGs is that we have a lack of available resources. Russ: Right. The IESG has discussed this numerous times -- about how many is too many? How many existing WGs could be eliminated? Mike: Do you have any idea what the load is for the IETF on a current 5 day schedule? Jari: It's been my observation that quite a few working groups that did not need all of the time that they requested. So we've been talking about the quality of time versus the quantity. Sean Turner: That's why the blue sheets have the start and end time, so that we can track time used. Unknown: New topic. Technology has improved so much and they are using 3D and color to convey things in schools. In different presentations whenever we use color it improves our ability to convey protocols and their interactions. But IETF coninues to use gray, dot dot dot pictures. Can we consider 3D and color pictures in RFCs? Russ: We have had this discussion many times, but we have put it on hold for RFC series until the transitional period finishes, hopefully in March. This will be a debate that the permanent RFC Series Editor is expected to lead. And the IETF Stream will certainly be represented in that discussion. Stewart: This is a controversial subject, and I've had people at this meeting express frustration at the constraints of ASCII art. But I remind everyone that you can publish non-ASCII version of RFCs; the requirements is simply that the non-ASCII version is not the normative version. David Harrington: I recently discovered that people did not know that the xml2rfc tools supports exactly what Stewart mentioned. You can put a pointer in the file to an image file which will show up in HTML and PDF versions. Russ: I see no one else at the microphone. Have a good evening.