______________________________________________________________________ DRAFT TRANSCRIPT Session: APNIC Member Meeting Date: Friday 3 March 2006 Time: 2.00pm Presentation: APNIC meeting program restructure Presenter: Randy Bush ______________________________________________________________________ PAUL WILSON: Randy has been one member of a discussion among SIG chairs about the possibility of adjusting the APNIC meeting program in future. That is the way that the SIGs are running through the schedule of the meeting, the location of operational content as opposed to policy content. And this is something which has been under discussion and I think we are going to hear a proposal for an experimental trial of a new format for the APNIC Open Policy Meeting, possibly coming up at the next APNIC meeting. So Randy. RANDY BUSH: Thanks. Again, I'm holding the bag for Philip. He's taken the lead on this. This is something, you know, he has one foot in APNIC and one foot in APRICOT and, as we saw this week, for instance, the routing SIG versus the routing track in APRICOT. These issues and also within the APNIC meeting itself, we find ourselves running between different SIGs because there are things of interest to us. And Philip suspects and puts forth the hypothesis that the reason is a large amount of content in some of the SIGs, especially the Routing SIG, but also some the DNS SIG and maybe some of the v6 SIG, etc, is really just general operational content of interest to everybody. And so what Philip proposed and we've been discussing among the SIG chairs - well, discussing at the SIG chairs meeting but discussing on the mailing list for a while - we're talking about for the next time when we meet, wherever that secret place is - I know it's on the planet Earth and probably in the Asia Pacific region - we would like to try an experiment. I am going to describe it and we're really looking for your comments and feedback on this experiment. And the first part of the experiment would be Monday would be a general meeting and let's call it APOPS because it's those things - sorry, Wednesday, pardon me - and it would be those things of general interest to the community, contributed from each of the SIGs, from Routing, from DNS, from v6, etc. And we would suggest trying, because this is operational content, so much of operational content is what's current in the last two weeks. And something that happened in NZNOG and is happening at NANOG is something they call 'lightning talks'. And that's maybe an hour or so at the end of the day for talks that are submitted within the last week or two. They're each 10 minutes. They may or may not have foils - give me the real term - Powerpoint - thank you - they may or may not have Powerpoint. Foils - that's a little bit old, huh? And of interest to the whole community again but less formal, less reviewed, get them in there and say something of interest to everybody, etc. Thursday would be everybody's interested in the Policy SIG so the first half of the day would be the Policy SIG. The second half of the day would be room for at least four SIGs but might not actually need them, for SIGs that have some very limited audience for the subject they want to stick in there. Like something, you know, to do with DNS Operations and how it interacts with the registry and, you know, 70% of people aren't interested. Or something obscure about routing, which might interest me and 10 other people. And so the SIGs - the second half of Thursday, Thursday after lunch, would be room for various SIGs to meet, OK? And then Friday would be as it is today - the general APNIC meeting. And that's the story. Please input and comments. PAUL WILSON: I think if I could add one aspect that was discussed and that is that this whole reorganisation - experimental - would need some sort of guiding program. RANDY BUSH: The program committee would be the SIG chairs, as it stands today, the same people put together the program, essentially, although this is not a closed cabal so anybody who wants to do work - you know, we never turn down work. The other thing that I forgot to mention is notice how well this would integrate with APRICOT a year from now. OK. APRICOT will also be organising a little differently into much fewer streaming tracks and this allows general sharing with APRICOT. So this would be probably a joint session or something like that. Totally undiscussed and un-worked-out but, instead of having the Routing SIG in APNIC opposite a routing track in APRICOT. Make sense? Bill, could you tell us who you are? BILL NORTON: Bill Norton. I work for Equinix. It always struck me as a little bizarre that one would have an IX-specific thing for APNIC and a separate IX thing in APRICOT. So I'd like that there is some combining of some of these things. It strikes me as really kind of funny that, at NANOG, we'd always been for years and years a single track where everyone gets in the same room and hears the same presentations and, if they don't like the presentations, they go out in the hallway. NANOG is in the process of switching to having multiple competing streams at the exact same time that RIPE is going from multiple tracks into having more of a single session. It's funny that we're doing the same exact thing here. We're exactly flip-flopping. RANDY BUSH: I don't think it's that strong. I think NANOG is trying to add some parallel, not all of it, just some parallel. We're taking what was all parallel and trying some single sequential and I think what's happening is everybody is moving towards a combined scheme. BILL NORTON: I think the strongest thing you can do along with this is to make sure there's some cooperation and coordination with the other APRICOT components and I suspect that's what you're doing. RANDY BUSH: This sets up for that and I believe - I don't really know because I'm not part of the APRICOT conspiracy. I have enough work to do - that that is indeed happening. I don't think it's actually started to be worked out but I believe both sides are putting pieces in place to be able to do that. Geoff. GEOFF HUSTON: I notice almost a subtle distinction - RANDY BUSH: Who are you, Geoff? GEOFF HUSTON: Geoff Huston, APNIC. I notice what you're saying is ops stuff and what the RIPE community said was open plenary. Now, they look much the same but, when you start label it 'ops stuff', you start saying, "I only want this kind of material versus other sorts of material." I suppose, in the SIG chairs' collective heads is this truly ops stuff or is there more of a characterisation of plenary material that would be of interest to anyone who attends an APNIC meeting? RANDY BUSH: Being over 60 years old, I know how to say, "I don't know." GEOFF HUSTON: Cool. Thank you. PAUL WILSON: Are there any more comments about this? I'd be interested to hear what some of the SIG participants, the other participants in the SIGs would think as well as some of the other SIG chairs. This is potentially a little more work for the SIG chairs if they're to be assembled into a combined program committee. We've never had that before and a combined program committee would also help, I think, to ensure some coordination and consistency between the different sessions. RANDY BUSH: And further, when we sat as SIG chairs to discuss this, it was thought that maybe the different subjects would take chunks of this - so there'd be the Routing SIG chairs would go out and specifically work on getting good - a few good routing presentations that are of interest to everybody - the DNS, similarly, etc. GEORGE MICHAELSON: I'm sort of channelling Joe Abley. Sanjaya and I found ourselves caught in having to be in rooms at SIGs but desperately wanting to be in another room and I know Joe would feel the same. I had no idea this was in planning and I had written to Anne and said, "Couldn't we consider a single-strand meeting?" And she said, "Watch this space." I think it's lovely. RANDY BUSH: It's not an all-or-zero experience. As human beings we can take ambiguity and modification. We can modify it before we try it, or try it and modify it afterwards or say, "This was horrible," and promise never to do it again. KENNY HUANG: I'm also interested in this kind of experiment but, if you're running this kind of exercise, I would recommend - because actually the issue was raised by Kazu-san and I recommend we provide guidance for the audience because they probably don't know whether they need to attend APOPS or other SIGs so we need to provide guidance to show which sessions they can participate. RANDY BUSH: The sessions aren't in conflict any more. It's not a choice. This is the APNIC meeting - first, second, third day. It's not in conflict. PAUL WILSON: There's parallel SIG sessions. KENNY HUANG: I mean for presentations. RANDY BUSH: Where their presentation would go? KENNY HUANG: Right. Because they probably don't know which special interest group they need to submit a proposal and, if we can provide some sort of guidance, they probably have a clear picture of how to submit a proposal. RANDY BUSH: Good point. Thank you. SERGE RADOVCIC: Probably an independent view but just a little word - one of the SIGs I go to is the IX SIG and I see smaller IXPs presenting and I'm worried they may be scared off in the plenary session by doing this. Will they lose out and not be of interest enough? RANDY BUSH: I think the thought is that the IX SIG probably stands, you know - there are two parts of the IX SIG, maybe, and I shouldn't be actually going to this space at all but I think, just as you split the IX SIG this week and there is the things about generic things that exchanges have learned and philosophy versus SIG reports, maybe that's what's going to happen here. But that's why the SIG chairs are on the program committee. And I'm sure, if you have a strong opinion and are willing to help, we wouldn't throw you out. SERGE RADOVCIC: OK. Thanks. RANDY BUSH: Geoff again. GEOFF HUSTON: If there's anyone else, I'll sit down and let them speak. RANDY BUSH: You're the only person standing. There's somebody standing in the back but he's not approaching the mic. GEOFF HUSTON: Geoff Huston, APNIC, once more. I'm not sure if I heard this quickly but one thing I noticed in the RIPE NCC and that kind of structure was that the SIGs were more oriented towards policy proposals - in other words to actually doing the work of policy and making sure that went through and the informational presentations, that seem to take up a fair deal of our agendas at this point, would tend more towards the ops-stuff blocking that you've got there. So that is, I think, one sort of difference to the way we currently operate now, that this would hopefully naturally bring out. What it does mean is that the interesting information is available to all of us and it becomes an issue of what particular policy proposals are sitting inside the SIGs, what's their status, where are they going, what do we need to do to advance them or whatever? It's my understanding that the SIGs focus on the policy development side of the work and the ops stuff is more informational and general interest. Is that a resonance with what you were thinking? RANDY BUSH: Yes, but it's worth noting that both at RIPE and ARIN the SIGs are a little more directly in the decision making of operations at the NIC. The database SIG in ARIN and in RIPE talk about policy choices at a very detailed level. I think that would be interesting if more of that happened here and this is trying to make room for that, to get the community more directly involved in the policy-specific aspects in the SIG and the general interest subjects. So that's the long answer. The short answer is yes. GEOFF HUSTON: OK. Thank you. It seems to me to be a good answer at this particular point in time. Try it. Yeah. RANDY BUSH: My answer to your last question was I didn't know. I felt I had to give you more meat this time. BILL NORTON: You call this an experiment. How will you know if the experiment is a success? RANDY BUSH: Like NANOG, this group has survey forms. But it's also let's see what presentations we get. Let's see how people like it, let's see how many people filled the room and let's see what they have to say afterwards. And, in this general member meeting, I think we're going to be standing up here again saying, "How did you like it? What would you like different?" But, if you have other means of getting a quantitative result, that would be interesting. KUO-WEI WU: I am Kuo-Wei Wu - RANDY BUSH: Let Bill just finish. BILL NORTON: Survey forms are great if you can get people to fill them out. This meeting is better because those who care about something can stand up to the microphone and say, "This thing was great," or, "This thing wasn't so hot." It would be useful to identify the specific problems you solve or specific goals that you're reaching for with this new format because then you can make the list of metrics that you can measure them against. For example, if one of the things you want to do was to get more speakers from a particular part of the world to talk and if you had this particular structure, you feel like you can get more people to speak at this topic, then you have the specific thing you'd be looking to measure. "Did we get more talks from Africa?" or one part of the world. But generally, I like what you're putting forward and I think it's probably the right direction to go. But it would be useful for people to know after the fact that we did in fact achieve something useful. RANDY BUSH: Point taken. KUO-WEI WU: Kuo-Wei Wu, APNIC EC. I think first of all, if this is regarding the APNIC open meeting with APRICOT, I think that is another consideration. If the APNIC is its own meeting for the Open Policy Meeting, I think first of all, our concern is that this meeting is really open for the membership to discussion about open policy. This is first priority. And the second point I think for the detail of the program design, I would prefer to go through the process, the regular process, maybe the SIG chair, the committee, you can go on to discuss what is a better way to organise it. I have no preference. But I will say the SIG chair, - the SIG chairs can sit together and to think about what is the best way. I have no comment. RANDY BUSH: The SIG chairs have been discussing this proposal for a year and then we had discussed it at lunch in Hanoi and then Philip's put it out on the mailing list a couple of times. It was not - 'discussing' is a poor term because there was not a lot of back and forth discussion although there was at the most recent SIG chair meeting and I think the SIG chairs were, in general, pretty happy with doing this. PAUL WILSON: Is there any comment from any of the other SIG chairs? Izumi? Che-Hoo? Kenny? Kazu? My suggestion would be to implement the program committee idea straight away and to ask all the SIG chairs to sit on that, on that committee on the mailing list and to discuss in a little bit more detail what this experiment would look like if it goes ahead. I think it's a valuable thing to have, in any case, a coordination body among the SIG chairs and you can put some more thought into this. OK. Thanks very much, Randy. And we have done all of the SIGs.