IETF 83 Administrative Plenary Minutes Paris, France Wednesday, 28 March 2012 Minutes by Mirjam Kuehne 1. Welcome Russ Housley welcomed the audience to the administrative plenary. 2. Host Presentation (Robert Pepper, Cisco) http://www.ietf.org/proceedings/83/slides/ slides-83-iesg-12-ietf-operations-and-administration-plenary.pdf Robert used to be a regulator; was at the FCC for 19 years before he came to Cisco. Member of the Strategy and Planning Group at the FCC. They made a conscious decision not to regulate the Internet. He thanked the IETF, because without the IETF and the rest of the technical community we wouldn't be where we are today. But we also have to realise that this place is threatened. If the Internet would be an economy, it would rank among the top 5 economies worldwide. The Internet was not an accident, and "unregulation" of the Internet was not an accident. There are real challenges to the Internet. Internet needs to be extended to everyone (to remote places, but also people with disabilities). We need to prevent illegal and criminal behavior and increase security on the Internet. Despite all these threats and challenges, the fundamental model and the approach how to develop the Internet don't have to change. Call for action: - Don't create a vacuum - Address issues before they become problems - We also have to talk to the non-technical people (business, legal and so on). They need to understand how the Internet works and what the best solutions are to solve the open issues. - We need to ensure the freedom of the Internet. 3. Reporting Russ told the audience that the time of this meeting was approaching and we had no host. Russ sent a mail to the IETF list and said that there was no host. David Ward responded and offered to help, and then within 5 days there was a signed MoU with Cisco. Russ thanked Dave and handed him a shiny green cape with the word "Superhost" printed on the back :-) 3.1. IETF Chair Report http://www.ietf.org/proceedings/83/slides/ slides-83-iesg-2-ietf-operations-and-administration-plenary.ppt Russ presented two proposals and asked the audience to provide feedback during the open mic session: - introduce a 'Bits and Bites' session similar to the 'Beer & Gear' that is taking place during NANOG meetings. This could provide additional sponsors for the meeting and would allow them to present themselves to the participants. And there would be free drinks and food. - remove the email addresses from the blue attendance sheets, scan them and add them to the proceedings so that they are accessible online. That would also make it easier to find the blue sheets if required by a subpoena. It also eliminates the need to store more paper. 3.2. NOC Report http://www.ietf.org/proceedings/83/slides/ slides-83-iesg-3-ietf-operations-and-administration-plenary.pptx http://www.ietf.org/proceedings/83/slides/ slides-83-iesg-11-ietf-operations-and-administration-plenary.pdf 3.3. IAOC Chair & IAD Reports http://www.ietf.org/proceedings/83/slides/ slides-83-iesg-4-ietf-operations-and-administration-plenary.pptx 3.4. Trust Chair Report http://www.ietf.org/proceedings/83/slides/ slides-83-iesg-5-ietf-operations-and-administration-plenary.ppt 3.5. NomCom Chair Report http://www.ietf.org/proceedings/83/slides/ slides-83-iesg-9-ietf-operations-and-administration-plenary.pdf Russ presented a plaque to Suresh to thank him for his hard work as NomCom Chair. 4. Open Internet Endowment Status Report http://www.ietf.org/proceedings/83/slides/ slides-83-iesg-10-ietf-operations-and-administration-plenary.pdf 5. Thank Outgoing IAB, IAOC, and IESG Members Each outgoing member received a plaque to thank them for their hard work. 6. Introduce Incoming IAOC Members 7. IAOC Open Mic Old and new IAOC members sit down on the stage. Cullen Jennings: Will the 'Bits and Bites' meetings stop us from having standards development meetings? That is, would the new event run in parallel with other meetings? Bob Hinden: No, we will try to find a time that is not in parallel with other meetings. Jared Mauch: Sometimes the IETF meetings are very close to NANOG meetings and also sometimes in the same location. Maybe we should consider some cross-pollination with NANOG? Adrian Farrel: I wonder what it would take to persuade the IAOC to run an IETF meeting in Africa? Bob Hinden: We would need a host. Also, don't forget we try to hold meetings in places where near many active IETF participants. Joel Jaeggli: Regarding the Bits and Bites meeting, Instead of concentrating only on commercial companies, maybe we could convince 'competing' standards organisations or other public interest organisations to present/market themselves during such Bits and Bites meetings. Some of them certainly have an interest in increasing the IETF awareness in their activities (for instance ICANN). Marc Blanchet: I hope that the IRTF logo issue will be settled soon. Bob Hinden: Yes, this will hopefully be solved tomorrow. Marc Blanchet: I would like to see the RFP on IMAP and mail archives be solved as soon as possible. Russ Housley: Alexey is working on an experiment using IMAP. What we learn from that experiment will decide the next steps for IMAP. Eliot Lear: I support the idea to scan the blue sheets. Lou Berger: I noticed that there was only one IETF protocol that could be used to reserve a room at the IETF hotel this time: email. Bob Hinden: I agree, and we were not too happy about this situation. Otherwise the venue is a good one. Charlie Perkins: I support the Bits and Bites idea. Regarding blue sheets, I like the idea of scanning them. What about the the email addresses? Russ Housley: The proposal includes removing the column for email addresses from the blue sheets altogether. Randy Bush: Can we improve the possibilities for people to participate remotely in WG sessions. So far, this doesn't work very well. Bob Hinden: Good feedback, thanks. Peter Saint-Andre: Would like to recognise the EFF and the others who stepped up and helped solve the timezone database issue (for background see Russ Housley's presentation). Yuri Demchenko: I suggest using a registration number along with the name on the blue sheet instead of the email address. Yaakov Stein: I noticed that there is a bar code on the back of the badge. That could be used for blue sheet purposes. Pete Resnick: Blue sheets are not used by us at all, except for the headcount and for answering subpoenas. And for that purpose we can decide for ourselves what type of information we want to collect. So, it is perfectly reasonable to keep the names if we think it is worth keeping that type of information. In Hiroshima, we did an experiment with RFID chips in the badges. It was nice to have the badges scanned at the microphones and then have the name and information displayed on the screen. Scott Bradner: The requirement to list participants is a very old one and predates subpoenas. As part of an open standard development process we need to know who participated. If we leave out information on the blue sheets, that might violate that requirement. He further mentioned that more remote participation would also pose that problems. If we want to go that direction, we will have to revisit that requirement. Paul Hoffman: Blue sheets are often unsigned or unreadable. If we have the rule Scott mentioned, we have been ignoring it for some time. The rule has to be revisited, but I suggest the blue sheets continue to be signed as before. In the meantime, even if the writing is not readable, we can scan them and publish them. But people have to realize that the blue sheets don't necessarily list all the participants of a WG. Scott Bradner: That is perfectly acceptable. Tony Hansen: Having the names is definitely be useful, the emails not so much. Joel Jaeggli: Another source of contributors that participate in WGs are those in the jabber rooms. If you participate in the jabber room, there is a record of that. Bob Hinden: Yes, and those records are included in the proceedings. Scott Bradner: Automating blue sheets came up before, also related to the RFID tag experiment in Hiroshima. So far, people have not been convinced this is useful or desirable. Christopher Liljenstolpe: I like the idea of a Bits and Bites session, but would be curious who would show up and pay for our beers. As someone who had to participate remotely from time to time, he thanked the jabber scribes for all the work they are doing. It would be nice, however, if the remote participants could be channeled better. It is irritating if your comment is presented 20 minutes after the point has been discussed in the room. Wesley George: I am not aware of documentation for jabber scribes. Maybe we need to provide some guidelines. John Klensin via jabber: If the only identification in the jabber log is a deliberately cryptic or over-cute handle, we actually don't know who was participating -- especially if the question arises a few years later. Lou Berger: We have to be careful when writing jabber guidelines; some people might just want to get a personal or marketing point into the meeting. John Klensin via jabber: Also, the jabber situation seems to be worse at this meeting because more attention is being paid to Etherpad. Tony Hanson: There might be some use cases for the email addresses on the blue sheets. Some things during a WG session are easier to do via Etherpad than jabber. Peter Saint-Andre: I know quite a bit about jabber, and we have capabilities to find out who is behind a nick name etc. If you want more jabber features, I am happy to help. Chris Liljenstolpe: Let's pick one method and stick with it: jabber, Etherpad, others? There is not enough screen real estate to keep track of all of them during a meeting. Bob Hinden: I agree that at least during one WG meeting, only one technology should be used. Thomas Narten: We have the need to understand who we are interacting with in the IETF. That becomes increasingly difficult, but it is relevant information. An IETF insider may be able to figure it out who is participating (attendees list, proceedings, etc.), but for an outsider this is hard to do so. And in the interest of transparency and openness, we might want to revisit that. Bjoern Zeeb: It would be helpful for newcomers if the acronyms used in the reports would be spelled out and if the role of all these bodies could be explained briefly before they report on their activities. Bob Hinden: That is a very good suggestion. [Then, he briefly explained the IAOC and the IETF Trust roles.] Randy: Jabber? Text based typing? Grow up! ;-) 8. Introduce Incoming IESG Members 9. IESG Open Mic Old and new IESG members sit down on the stage. Eliot Lear: If people want to participate in the timezone database discussion, there is a mailing list: . Lucy Lynch: There is a bad collision on Thursday between two or three different WGs. Can you please explain the scheduling algorithm? Russ Housley: During the IETF Chair report, I talked about laying out a new grid and schedule for each meeting in the future, instead of using the grid from the previous meeting. He also asks WG Chairs to include only conflicts that would prevent the WG session from being useful. I also want to acknowledge the work done every meeting by Wanda Lo from the Secretariat; she does an amazing job juggling all these schedule conflicts. Barry Leiba: I'm asking WG chairs to work with each other in case people have to present in more than one WG. Maybe they can be at the beginning of one agenda and the end of the other, thus avoiding the need for a conflict. Russ Housley: The database now has information about each session. There is an RFI out right now for IETF meeting slot scheduling. Adrian Farrel to the audience: Would you rather suffer conflicts or suffer meeting on Friday afternoon? [A hum showed a relatively even split; no consensus in the audience.] Dave Crocker: We seem to have a situation of congestion. How to get out of there? We can add more capacity, but there are limits. We can use the tracker! Some WGs are very productive, others are not so much. The WGs not getting their work done are creating conflicts for others. Barry Leiba: Good suggestion, but many of the non-productive WGs don't meet during the IETF meetings. So, they don't actually contribute to the problem. Chris Liljenstolpe: I remember times when the schedule was less full. How much time does extending the meeting into Friday morning and then Friday afternoon buy us before we talking about extending the meeting to Saturday morning. Are we pushing away the problem? Keith Drage: RAI could have had a better schedule this time. Russ Housley: For IETF 84, we are going to offer 30 minute WG slots. Keith Drage: WGs have to realise that they do not have to meet. Instead they can do more work between the meetings. He also mentioned that he never gets to attend the General Area meetings because they always conflict with WG meetings in the RAI Area. Joel Jaeggli: The v6ops WG seems to create a lot of conflicts. He is pleasantly surprised that there is so much interest in IPv6 operations. But he doesn't understand why many of the WGs on the conflict-lists are there. Russ Housley: I agree, and I urge people to review the conflict lists. Pete Resnick: To make most productive use of face-to-face meetings, WGs should talk about open issues. Reporting and preparation work can be done on the mailing list. Robert Sparks: Could we possibly make it easier to have WGs chairs see what other WGs list their WGs as conflicts? Marshall Eubanks: Another possibility would be to have another IETF meeting per year. Russ Housley: That was considered before. For instance there is a RIPE Meeting scheduled in Amsterdam in September 2012. We are considering an interim meeting there. Dave ?: For those WGs that have so many conflicts, maybe the scope is too big for the charter? Tony Hansen: There are groups that span more than one Area. Maybe they can meet in parallel? Possibly that could help reducing conflicts? Also, when we fill in the list of conflicts, maybe we should prioritize the conflicts? Harald Alvestrand: Maybe people are thinking too narrowly about the problem. Most of the conflicts are created by individuals who have to be at more than one place at the same time. He suggested to slice the agenda into 30 minutes slots and to assign these slots based on I-Ds and work items. Andrew Sullivan: Perhaps we should fill the agenda slots earlier so that people are forced to manage the time slots they get properly. Gonzalo Camarillo: The feedback the IESG received was that people want to work closer to the meetings, not earlier. Dan Romascanu: There are other conflicts that might not be as obvious. Shawn Emery: Some presentations in some WG meetings could be shortened a lot. John Klensin via jabber: Pete, would you and your colleagues be willing to ask WG Chairs at scheduling time what specific issues they anticipate needing to discuss at the meeting? Getting realistic answers might be hard given how early we start scheduling, but, if a WG has no clue, maybe that ranks no slot or at least a low priority for slots, slot length, or conflict avoidance. Pete Resnick: Yes, spot on. Similar to what Harald said: instead of scheduling based on a group, schedule based on issues or documents. Thanks for the advice, will think about it some more. Adrian Farrel: We have to react to the community and community seems to want 5 minute status update slots on the agenda. Marshall Eubanks: One of the consequences for introducing the 3-year meeting schedule is that we is less flexibility in changing the number of rooms and such at a venue. Peter Saint-Andre: We need to cut the traffic. Perhaps we're doing too much, perhaps we have too many WGs. Maybe we should be tougher when forming WGs that we don't believe will produce a result that will be implemented or deployed. Dave Crocker: There are many points of control to reduce the traffic. They are reasonable but challenging. We need to learn how to impose additional significant discipline. It is easy to say "this is what the community wants." But if we don't have the capacity, we need to come up with another solution. Chris Liljenstolpe: What is a face-to-face WG meeting meant for? Is it for 5 minute status updates or proposals that haven't been presented on the mail list before or is it to discuss open issues? Maybe things that haven't been discussed on the list, shouldn't get a speaking slot. Dave Harrington: We need to do both. We need to discuss open issues and present new ideas. Chris Liljenstolpe: But should these new ideas not be discussed on the mail list first? Dave Harrington: Yes! Lou Berger: Maybe it is a strange argument, but I am also sensitive that people have to justify travel budgets. Shouldn't take that into account? Gonzalo Camarillo: There is a tradeoff. We need to balance the time in the WG. It is strange that we talk about travel budgets. Randy Bush: It seems every possible alternative in a protocol is now written down and documented. Doesn't believe this is useful. "We have Internet Engineering Steering Group. Where is the Internet Engineering Breaking Committee?" Tim Shepard: I wonder if it would be an idea to have an opening and a finishing plenary? Or possibly combining the plenaries? Unknown: I considered presenting a document at the IDR WG that would have required a short status update. I decided not to do it. Instead I submitted a proposal to the WG chairs, and then it was submitted and discussed on the mail list. Suggest others do that more often. Marc Blanchet: This week, many documents are on hold because of a possible new WG. What happened to the idea of an IPv4-exit WG? Currently, this situation is stopping the work to be done. Ralph Droms: I apologise that this is taking so long. It is actually not trivial, and it involves many sensitive issues. Ron Bonica, Dave Harrington, and himself are talking about it this week. Hopefully there will be a plan, and some rules of what should be discussed in this WG by the end of the week. Dave Harrington: It was decided to close the BEHAVE WG. At the same time an IPv4-exit WG has been discussed we were holding off to decide how to recharter BEHAVE. In the meantime, CGN is growing, so we're not quite done with NATs. That makes the decision about IPv4-exit difficult. And we are looking for volunteers to run this WG. Peter Saint-Andre: If we ask the WG chairs to manage their time slots more efficiently, then maybe we should do the same in the plenary. Let's get rid of the reports and create more open mic time and discuss the open issues. Russ Housley: A certain amount of reporting is necessary. Peter Saint-Andre: Post the slides to the list, and then ask people to read them before the meeting. Eric Rescorla: The reporting part of the plenary should not take more than 10 minutes. It should be basically a list of links. Russ Housley: I will take an action to see what we can trim on the reporting at the administrative plenary.