IETF 81 Administrative Plenary Quebec City, Canada Wednesday, 27 July 2011 Minutes by Dean Willis 1. Welcome Russ welcomes the IETF participants and presents the agenda. 2. Host Presentation RIM expressed their gratitude expressed to Their Royal Highnesses, the Duke and Duchess of Cambridge for visiting the city in advance of the meeting and helping to prepare it. Presentation reviewed role of Quebec as the firewall manager for the 18th century trade network (see slides). The IETF participants thank RIM for hosting IETF 81 and Russ presents a plaque to RIM. 3. Postel Award Postponed The Jonathan B. Postel Award is traditionally presented during the July IETF meeting. The honoree of the 2011 award is unable to attend IETF 81, so the award will be presented at a time and place where the person can be present (see slides). 4. Reporting 4.1. IETF Chair Report by Russ Housley (see slides) Reviewed attendance at current meeting, formation and closure of working groups since last meeting, numbers of I-Ds new and revised, IETF Last Calls, I-D approvals, other requests, and RFCs published. IANA performance has continued to be good, with substantial progress on format conversions. RFC Editor is in the process of setting up digital signatures for RFC series documents and has made RFC 1150 historic and concluded the FYI sub-series. There have been a number of late IPR disclosures, such as the one regarding RFC 6073. Some are calling for penalties. Code sprint was successful, including datatracker and xml2rfc revisions. Future meeting calendar reviewed. Noted that a host is needed for IETF 83 in Paris. 4.2. NOC Report by Noah Weis (see slides) Network has links from Telus and Bell CA, both 1Gb/s and covers the Hilton hotel wired network. Multiple SSIDs supported on wireless. Traffic peaking at over 100Mbps. Wireless associations graph is showing new usage patterns due to people with multiple devices, with even balance between 2.4 and 5GHz connections. Noah thanked many volunteers (see slides for list). 4.3. IAOC Chair and IAD Report by Bob Hinden and Ray Pelletier (see slides) Bob Hinden spoke first. Almost caught up on minutes; 9 of 13 published. Need more volunteer scribes, otherwise professional aid will be needed. The Prague meeting had good attendance, with a net of $502K, which is $30k above plan. Quebec attendance is slightly below estimate. 2011-2014 budget plan introduced (see slides). Detailed budget for 2012 will be developed over next few months. Financial history from 2006 reviewed. RFC Production Center and RFC Publisher contract being extended for two years, and RFC Publisher contract realigned on calendar with Secretariat contract. The ISE contract is also going t be either extended for 3 years or a new search will occur. Status of advance 3-year advance meeting planning was provided, and 6 of 9 meeting venues are contracted or under negotiation. Asia venues have become challenging, with room rates over $300/night. This may make Asia too expensive for yearly meetings. Many future meetings need hosts. Much progress made on IETF tools. Thanks to Henrik Levkowitz for his contributions as program manager.    Ray Pelletier spoke second. Acknowledgement of staff and volunteers. Tools development acknowledged. Highlighted major improvements to the Datatracker to include the whole document life-cycle. IAOC selected winning contractor for back-end development today. Thanks given to RIM as host of this meeting, as well as Bell and Telus as circuit providers. 4.4. Trust Chair Report by Bob Hinden for Marshall Eubanks (see slides) Role of the IETF Trust reviewed. It is now 5 years old. Thanks given to CNRI and ISOC, the two organizations that made contributions to create the IETF Trust. Trust provisions have not changed since 2009. Thanked scribes for making minutes available, and we need more volunteer scribes are needed. IETF Trust occasionally receives subpoenas including requests for documents and proof of attendance. Three have been received so far this year. 4.5. NomCom Chair Report by Suresh Krishnan (see slides) Suresh is the NomCom Chair this year. 120 people volunteered to be voting members this year. The 10 voting members that were selected were announced 26 July 2011. The focus for this meeting is self-organization and gathering of feedback. Membership was asked to stand up and be introduced. Open positions are presented in slides. Nominations are requested, and feedback to NomCom is urgently requested. Noted that discussions with Nomcom are confidential. 5. Recognition Marcia Beaulieu has been supporting the IETF for 15 years. She has been part of the IETF Secretariat at CNRI, Fortec, Neustar, and AMS. Acknowledged here contributions to with flowers, a crystal, and a travel assistance gift. The audience made an extended standing ovation. 6. IAOC Open Mic Dave Crocker: Marshall Eubanks, are you saying that there is an opportunity for people to contribute to the IETF by volunteering to be scribes? Marshall: YES! Paul ? : Are subpoenas public? Should they be? Marshall. That's an advice-of-counsel question... Jorge Contreras: There's been no need to keep them secret. They could be made public, and we could post PDFs if desired. Barry Leiba: Recommends volunteering as a scribe; it teaches how the organizations work and gives social interaction with the people. He summarizes needed skills as being able to write fast, understand cross-conversations, good English. There are different styles, from literals to notes. but all of them help the organizations remember and review. Charlie Perkins: Short question about scribing. Can you ask questions as a scribe? Marshall: Yes, if something is not clear, you can. Charlie Perkins: The network in my hotel is bad enough that I would have changed hotels if I had known. Is there something we can do about this. Bob Hinden: We typically take over the main hotel network, but not the others. We try to vet the secondary hotels, but mostly they have not-quite adequate bandwidth and there's not a lot we can do about it. Perhaps we should recommend IETFers not abuse their hotel networks. Perhaps we can publish information on the hotel bandwidth in the future. Spencer Dawkins: Volunteered as IESG Scribe for 4 years and recommends it. Notes that the job is much safer since EKR is not on IAB (humor). Glen Zorn: Amazed that Asian hotels are so expensive. Asserts that there must be something wrong with search criteria. For example, it is hard to find a hotel in Bangkok that costs $300 a night. Bob Hinden: The criteria also include airfare and target dates, and that the inflation of the dollar has also been a problem. It sounds easy, but the people who do this for a living are finding it to be difficult. Noted that there are usually alternate hotels, but that at some venues the alternates were not affordable. Glen Zorn: About hotels. Asked on list and didn't get a good answer. Up until about 3 weeks ago it was possible to book at Hilton for a lot less than the IETF rate, like $100 a night difference. Why? Was it changes over a 3-year basis and futures? Bob Hinden: Many factors go into the negotiation, and sometimes we win on rates and sometimes we lose on rates. Glen Zorn: Clarifies that he attends IEEE meetings, and they have a big problem. They're having to extend a discount on meeting fees to get people to stay in their hotel blocks, because people just won't stay there. Bob Hinden: Could we try to get a thousand rooms on Priceline and see what it happens? Dave Crocker: Hotels are arguing they're offering better rates or better rooms or something. Glen Zorn: Until recently, it was uncommon to find cheaper-than-block rates in the rooms for IETF meetings. Ole Jacobsen: Perhaps we can put contractual protections into place. Perhaps we could contract against a fraction of the rack rate. Glen Zorn: We're not taking sufficient advantage of our business. It might be possible to balance subsidy between hotel rate and conference rate. Bob Hinden: We have tried balancing games here, but have to be careful. Phillip Hallam-Baker: Have we considered places like Egypt and Tunisia, and looked for sponsors in the social networking world? Bob Hinden: We look for sponsors from all Internet-related business. Paul Kyzivat: Recalled lack of cookies in Paris. Here we are five years later, short on cookies again. What is it with French-speaking countries? Can we get somebody to look into the cookie situation? Cullen Jennings: (said nothing but gave Paul a cookie) Andrew Sullivan: If you scribe for the IAOC, you'll get to learn about hotel negotiation. IAOC thanks Andrew for his scribing. 9. IESG Open Mic Kevin Fall: I need to know what to call the 2nd byte of the IPv4 header. RFCs 3260 and 6145 disagree. Answers include "The byte formerly known as TOS", "The second byte", or "The second octet". Glen Zorn: IESG has as a group over the years has shown itself to be basically incompetent in the review of drafts, perhaps throwing a DISCUSS when they don't understand a draft outside their area. They're really way too busy to learn this stuff. He would like to suggest that IESG members recuse themselves under these conditions. But as this is unlikely to happen, he suggests that IESG should cease to provide mandatory technical review, but instead, when appropriate, IESG members would provide individual review where appropriate. We've got all these directorates that should be doing the technical reviews. David Harrington: likes to think he's providing a useful service by making sure that drafts are readable at the level of a beginning programmer. Glen Zorn: A DISCUSS if you can't read it is okay, but a DISCUSS because you lack the technical expertise to evaluate is not fair. Nor is it fair that we expect every IESG member be able to evaluate every last draft. Jari Arkko: This is to a certain extent happening now. Yes, ADs do occasionally issue inappropriate DISCUSSes, but mostly they push back on each other. Ralph Droms: IESG is primarily providing quality control. This could be done somewhere else. The IESG may be finding too many issues too late, but we can't go without any quality control. There may be a solution, but it is not as simple as just not doing quality control at the IESG level. Randall Gellens: IESG review can be helpful, but as Glen mentioned, inappropriate DISCUSSes are a problem. This includes differences in style or technical choice between AD and WG. They really should be a "last resort". The COMMENT process should be used for most pushback; DISCSS should be reserved for serious problems. Russ Housley: The IESG criteria document for a DISCUSS, and authors and WG chairs should help us enforce it. Pete Resnick: The document shepherd and other responsible parties really should push back on the AD. Peter Saint-Andre: DISCUSS means "Let's have a discussion". Dan Romascanu: Directorates were formed to do earlier review. Jari: His tracking tools show no real change in DISCUSS probability over time. He is of the opinion that ADs should perhaps DISCUSS less. Pete Resnick: He has been trying, with limited success, to get the Apps Review Team involved earlier in WG process. They don't have to wait until Last Call. Sam Weiler: For years, the NomCom has been reporting that it is hard to get people to volunteer as IESG. Perhaps changing the review model to reduce the load at the IESG level would make this better. Phillip Hallam-Baker: What are we going to do about the transition to two maturity levels, or how am I going to get HTTP to full standard? Russ: As author of the two-maturity-level document, it seems to be proceeding. It just entered IETF Last Call today. Jari: Thinks the new proposal is good. He is also frustrated that process is so hard to change, and wants small steps for improvement. But if the community doesn't like the two-maturity-level proposal, it doesn't go forward. Keith Drage: I'm one of those who would like to streamline. It does not matter that much if the IETF does occasionally publish crap. The authors names stay on it. Notes that shepherd review seems to have been effective, and it would be good to have statistics on individual shepherds performance with respite to IESG review. Russ: It is IESG's job to ensure that cross-area review takes place. Gonzalo: Many shepherd and working groups don't really know what to do with a directorate review. Henning has shown work that indicates there may be a point of diminishing return on reviews. Elwyn Davies: Has been on Gen-ART review team for years, and has seen few frivolous DISCUSSes. He would NOT like to see the IESG become totally a managerial exercise and likes them to be technical gatekeepers. (Substantial applause noted.) What damage does a frivolous DISCUSS do? Barry Leiba: DISCUSS means have a discussion. This is fine unless the AD is unresponsive. He's had to explain stuff to ADs and feels okay with it, but gets frustrated when the discussion can't happen. Keith Moore: Notes that over twenty years, there aren't enough cookies, the net is too slow, and the IESG is not doing its job. People who don't know what they're talking about like to criticize those who do. We now have more tools and transparency than ever before, and encourages ADs not to abuse DISCUSS, but use them when needed. Pete Resnick: It is inevitable due to volume that ADs will end up DISCUSSing things they don't need to. (New AD syndrome?) and welcomes ongoing pushback from the WGs to help keep things working. Harald Alvestrand: When Glen raised this question, I had to check my watch to remember which decade we are in. Really, we've improved the process quite a lot, but that incremental process will not change the question and we'll have this discussion in 2215. Thinks the IETF needs to break out of the mold of this process, even though it works fairly well. Applauds the two-maturity-level work as the largest piece of reform since forever. But we need more. Gonzalo: We're doing some cross IETF-W3C collaboration work, and that's a BIG change. Randall Gellens: The shorter the document, he more silly DISCUSSes we get. Jari: This is confirmed by statistics. David Black: Thinks the directorates are making a real difference.