Carsten Bormann (T2TRG Chair)
Jenny Bui (IETF Secretariat)
Ignacio Castro (RASPRG Chair)
Stephen Farrell (UFMRG Chair)
Liz Flynn (IETF Secretariat)
Mirja Kühlewind (MAPRG Chair)
Dirk Kutscher (ICNRG/DINRG Chair)
Colin Perkins (IRTF Chair)
Alvaro Retana (RASPRG Chair)
Ali Rezaki (Sustain Chair)
Eve Schooler (Sustain Chair)
Michael Welzl (Sustain Chair)
Lixia Zhang (DINRG Chair)
Jenny: Does anyone have anything to add to today's agenda?
Colin: I have two very small AOB things at the end, but nothing major.
Jenny: Does anyone have any objections to the minutes from the November 26, 2024 IRSG Teleconfernce being approved? It was quite a bit ago if anyone remembers that teleconference.
Colin: I think I reviewed them at the time and I think they look good. So as you say, it's been a while.
Colin: So all we need to do there is get the text and copy it to the Wiki somewhere cause it currently lives in Melinda's GitHub, which is maybe not the best place for it.
Jenny: I can do that, if you want me to.
Colin: If you can do that, that would be handy yeah because we will just, we will just forget where if it's there cause Melinda's not coming to these meetings anymore.
Jenny: Ok. I will do that and add it to the Wiki.
None
None
Colin: I don't have much to say about those. We have opaque and aead-properties if I am reading it correctly in the RFC editor queue and I think they're both fine. It's progressing smoothly.
Colin: Yes, we got one out, so thank you Mirja, who dropped the ballot in. Let me take one document out of the final poll like half an hour ago. So I think the only thing left in there is the Code of Conduct, which we'll be talking about later today.
Colin: So we have draft-irtf-cfrg-kangarootwelve. We have draft-irtf-coinrg-use-cases and we have the draft-irtf-nmrg-green-ps, if I remember right. The first two of those I think are going fine, looking at the comments in the Datatracker for thedraft-irtf-nmrg-green-ps, it's on the IESG agenda for I think next week. The initial outcome unless the IESG changes their mind before next week is that there are concerns with overlap with the Green working group. So we may have to revise that document to bring it back and make sure it's not overlapping with what Green is doing or at least clarify the relationship between the two. But I guess we'll see once the IESG has their call next week. So I think we're under control that one might need a little bit more work. And if you look in the documents in the Datatracker, there's an IESG conflict review and there are some comments in there. That will come back in a couple of weeks I think.
None
Colin: So we've got the draft-irtf-iccrg-rledbat draft, which Mirja let us finish by balloting just earlier today. That one I think is ready to go. There were some nits that Brian found and I will check with the authors, I don't think they're significant but they might want to tidy them up but that those are editorial things and then once that's done, it will be ready to go to the IESG. The other one is the draft-fluhrer-lms-more-parm-sets, which is a CFRG document even though it's not draft-irtf-cfrg. The IESG were happy with it. There was a comment from IANA that needs to be resolved and the authors did send a follow up to me with a proposed response and that's just waiting for me to follow-up on that with IANA. So it's blocked, but I think it's under control and we know why it's blocked. I don't think it's anything difficult, it's just tidying up the registries and the IANA considerations.
Colin: We chartered the SUSTAINRG with Ali, Eve, and Michael as the chairs and I see them on the call. I think the group is planning to meet for the first time in Bangkok. Thank you everyone for the feedback and the review of the charter for the past few months. I suggested that the chairs might want to say something about what their plans are for the group and just introduce themselves and the group but so yeah over to you.
Eve: I'm Eve Schooler and I guess I've been involved in IETF for a long time, and I've been a Research Group Chair, I guess as far back as before MMUSIC started, I started something called CONFCTRL which started as a Research Group and then eventually graduated into a Working Group, MMUSIC, which lived for a very long time. Long enough to see me return to the IETF after a long hiatus, and then about three years ago, joined Marie-Jose and Jeffrey to help with the COINRG Research Group. And, now the Sustainability Research Group and so I'm really pleased to see this emerge as a topic that is really near and dear to my heart. I'm presently a visiting professor of sustainable computing at Oxford after a long time career at Intel. So that's me.
Michael: Been doing IRTF things for a long time and I've been away for a while from chairing. I've been using ICCRG as an excuse to never again have to chair, but I think I have to give it up at some point. I chaired this for eleven years, so here I am back doing IRTF and also, just like Eve said, the same for me, sustainability is pretty close to my heart. I think it's an interesting topic. It's exciting because it's different than people may expect it to be and it's harder to pull off. And so I'm expecting this group to bring some light into this difficult topic.
Ali: I am currently the head of sustainability systems research at Nokia Standards and before that I was working on security standardization, managing the team on 3GPP and other standardization bodies and for some years now I have been attending IETF and IRTF trying to, let's say, promote the sustainability perspectives that we have been working on and what it means for networking and I'm really happy that we are at this stage have launched sustain RG where in parallel, of course there is the standardization track going on in Green Working Group in IETF. We also have the e-impact IAB program, where we are also discussing architecture related topics. I have a 35 year career in tele-communications, sometimes in research organizations, sometimes in development of routers and switches. Starting from the early days of frame relay and going up to IP MPLS and eight going through ATM stuff, so quite aware of the challenges of implementation and operation as well. So I'm really excited to work on the SUSTAIN RG with Eve and Michael. Let me start briefly, and then Michael, you can also add what we want to do in Sustain RG. We want to be a place of discussion and advancement on the long term research questions related to internet and sustainability, starting first with the characterization of the impact of internet on environmental, social and economic sustainability, so mapping what the current status is and then the next challenge would be how to improve the status quo with respect to certain sustainability targets.Our immediate challenge is to engage the IETF, IRTF community, get your favorite topics into this, research group and advance on those. Of course we are aware that sustainability is an extremely broad topic, so we will try to define a vision and try to create focus. But our initial perspective would be to attract as broad an audience or contributors as possible. So, I look forward to it and our charter is published, our Wiki. We are trying to set up and hopefully we'll get going with our first meeting in Bangkok.
Michael: I can say a few things, maybe about that idea that I proposed in our last meeting when we talked about this. So this is not an exactly original thing coming from me, but, my background is only in ICCRG. The thing is that this group was created back then to take over agenda time from groups that were busy and too busy to look at new congestion control proposals. So we were actually created back then so that people would have a place. It was called throwing it over the fence so that it could be taken over and baked for a while and then come back. I thought maybe we could apply something similar here so that if a group has something to do related to sustainability and it becomes an issue, I mean let's take to, to be more concrete just the time-variant routing group as a topic. If they would drift off into talking about the sustainability use case longer and too much and that would start spending too much time on the agenda, but it is interesting, they could bring it over to our SUSTAIN RG, discuss it a bit with us and then maybe we could give them some reviews and some feedback and opinions and they could come back with it. Maybe we could try and apply the same model there for things related to sustainability in general across the IETF. That was an idea.
Eve: To give a little bit more flavor, some of the other things that we have talked about or bandied about in terms of where to head, directionally, it makes a lot of sense to liaise with other SDOs that are in this space because of course no networking layer exists in isolation. We're very connected to the rest of the internet architecture and so liaising with others SDOs is sort of the beginnings of bootstrapping ourselves. Additionally, how to plug into the already existing, very interesting Applied Network Research Prize and workshop, in hopes of offering our services to review papers that come in, possibly even try to encourage a track on sustainability. That's something that we've thought about, and along those same lines to and expand upon what Michael said was to view the group as a set of experts who could review for the broader IETF, IRTF community as these sustainability related topics arise in documents and so forth. We kind of see ourselves as needing to begin to map out the landscape, so like what's the navigational map of existing IETF and IRTF sustainability efforts like a snapshot of that and helping people sort of understand where we are, and some of that may involve really trying to lay down or articulate the terminology that's being used in this space. So you can hear there are many topics that even just to get started, should keep us busy for quite some time. Back to what Michael was saying with regards to other IRSG groups that we either have participated in or led or admired, so often the topics come organically from the groups themselves, from the people who are enthusiastic about various topics and I think we plan to embrace that philosophy here as well. Including our own topics that we're enthusiastic about.
Colin: But finding the set of people and herding the cats is a big part of it. I mean, I think that there's a bunch of interesting topics here and lots of different ways this group could end up going. I'm really quite excited about trying to get a track in ANRW going. I think that would be a really nice thing and would connect it really well with the research community. I I think your biggest challenge is likely to be keeping the work in a manageable scope rather than trying to do all of the things all of the time. I suspect I suspect you're not gonna need one of the small rooms, I think you're gonna need quite a lot of, you're gonna be one of the big groups at least for the first few meetings.
Eve: Honestly, I think it would be really great if this community, Colin and Dirk, when he begins, and others in the IRSG, could weigh in and help us along those lines. We welcome your feedback and guidance with scope because we recognize it could be completely everything, you know, boil the ocean, but we really don't want that. And so please feel free to pipe up.
Colin: Absolutely, and if it's useful to have a call to talk about how to run things and do some planning and I'm sure we can arrange.
Ali: Thank you so much. We'll take you up on that one.
Colin: Yeah, looking forward to it. Tick the box for one of the big rooms when you request the session.
Eve: Well, in addition to guidance, think about the topics that are top of mind for you and your work or in your leadership in your groups. We've just sent out the solicitation for the upcoming meeting for topics and talks.
Colin: We've been talking about the Code of Conduct on and off for, I guess a year or 18 months or so now, and this is draft-perkins-irtf-code-of-conduct. We had some discussion back in the meeting in November just after the last IETF and mostly talking then and on the list about the text around use of generative AI. I revised the draft I guess a week or two back earlier.
This month to update that text based on that discussion and submitted a new version which just changed that one paragraph to align with that discussion. I think this addresses all or at least most of the concerns in the comments on that text from Mirja and from Jana. Jana certainly has said he is happy with it. I think it has made at least some of the changes Mirja was talking about. I think the only open issue I'm aware of with the draft is that Mirja said she thought that perhaps the right thing to do would be to remove that paragraph entirely I'm sure she's on the call and can correct me if I'm mischaracterizing her view, I didn't think it was perhaps settled as to how things were playing out in that space. My feeling is that we actually need this paragraph, I think if we remove it, we're quite far out to step with the way other organizations such as ACM are going in their code of conduct and policies around generative AI. I also think until we know how societal norms around use of generative AI are going and until there's a sort of really settled view on how the use of AI generated text affects copyright and passentibility and things like that. I think we do need to err on the side of caution and request that people disclose.
when they've used AI to help generate ideas and and you know this is about generation of ideas rather than using AI to create grammar and topics like that. So, I think what's in there is the right thing to have and I am suggesting we move forward with the documents as it is. But if Mirja is on the call or if anyone else has opinions I would certainly like to hear feedback and comments.
Mirja: I actually balloted yes or already balloted yes, before and I just updated my ballot, I think it's fine to publish as this. The reason why I suggest to remove it is because we feel like it's a moving target. It's like we don't really know what the right thing to say is and kind of the code of conduct is a little bit written in stone, right? I mean of course we can update it, but I think it should be in a level that it's actually valid for a long time. And I might also think that on the long term because all these AI tools are now everywhere and which is like sometimes hidden and we use them like more on a daily basis more and more. It might be enough just to apply the normal author rules here like if it wasn't your idea you should not attribute it yourself, right? That's just the normal thing. And so I'm not sure we really need to say something in detail, but I'm also ok with the text. I mean, again, it's supposed to be long lived, but if we figure out we need to have a change in the future we can do it, of course.
Colin: I guess the authorship thing I think gets a little bit tricky because if you're not crediting yourself with authorship, and you're the author and you're not acknowledging the AI, well, who are you crediting? I do see what you mean though. That things may change here real relatively quickly.
Mirja: I mean that was also the thing that I'm really not sure about what is the right thing to do because, you know, if I read a website and they have a good idea and based on that idea, I develop my own idea which is slightly different. I wouldn't attribute the website today, right?
Colin: I would say. You should if.
Mirja: It depends on what exactly the idea is, right? If it's really related work, if it's base work, I should, right? And maybe that's one of the challenges when using AI that you don't really know where the ideas come from, so you cannot attribute the original source. At least and I mean for ACM I think it's more OK they reacted very quickly. They have this policy already for a year or two or maybe longer, I don't know. And they can just modify it and I think it's very useful to offer some guidance in the space and maybe we could definitely write something down but I think also like over the last one or two years, our thinking has changed a little bit already.
Colin: Do other people have views?
Dirk: I share the concern I mean this might look quite outdated in just a few years.
Colin: Yeah, it might end up being outdated.
Dirk: The first sentence definitely still makes sense so it must not be listed as authors.
Mirja: I mean even that one, you don't need to necessarily write in a code of conduct policy. I mean it's at the end, it's the RFC Editor who should enforce that.
Well, but I mean, again, from my point of view, I did vote yes already and if it's really outdated at some point we need to update, that's it.
Colin: I mean we can update these things if we need to. I suspect given all the discussion of email moderations and everything in the IETF is likely we'll end up updating it to talk about the changes in the way mailing list moderation and so on happens in a couple of years as well. Because if that starts working and there's a sort of community wide moderation team, I suspect the IRTF will end up opting into that just to avoid diverging too much.
Dirk: But it may be ok. I mean, there may be other updates that we learn about anyway, so I guess this will be updated just based on your experience, so could just apply the same approach to this paragraph?
Mirja: I mean maybe the other approach we could take is, to say slightly less and just kind of raise the topic and then say that author should align with normal author practice in the publishing world when e.g. site the ACM policy as an example.
Eve: That would certainly free us up from language getting old.
Dirk: Yeah, unless we have any special requirements which I don't see, so it could be fine to refer to it like an existing policy.
Colin: I'm a little cautious about that cause then that policy ends up changing and we don't realize it’s changed. I mean we could also just weaken it slightly and say should be disclosed or something which has a bit of wiggle room norms evolved for people to say well it's not a very significant use or whatever.
Mirja: That was another thing. You could add the word significantly here and then it's a little bit more judgment code of the authors.
Colin: Use generative AI to create significant amounts of texts rather than just to create text. That might be, that might be a way of doing it.
Dirk: Maybe add that qualification and then you can remove parts of that paragraph like so that refers to spelling and grammar checks and so on you don't really need that anymore.
Colin: Probably leave that in because that's clear, it's not a problem if you just do those things, it's more that it's generating new stuff, which is an issue. That's the case where it's a click, if you just fix grammar then it’s OK.
Dirk: Right. Not sure that that is. I mean it's really needed, I mean because people are just using editors and there they have these features already.
Colin: They do and that's why I wanted to sort of highlight it because I don't want someone freaking out and saying I've used this editor and it does this stuff and, now need to to sort of credit the fact that I use text editor or whatever.
Mirja: That was kind of my point. I think right now it's useful to have the sentence but it might might very well be very quickly outdated because AI is so everywhere, you might not even noticed.
Colin: Well, you're certainly not gonna notice. I can believe people will not notice if it's being used for just tidying up the writing I think I I have this conversation with the AI to brainstorm ideas, people will know they do.
Mirja: But I mean even these chatbots, they might be much more deeply integrated and all kinds of editing tools and whatever that you're kind of less aware of what happens in the background in future. But I think that part is not problematic. The problematic part is about the very strict requirement to disclose something, which might be out there very quickly.
Colin: Maybe change create text or other content to create significant amounts of text or other content and strike the last sentence if you're uncertain about the need you should disclose.
Dirk: I think it's ok.
Colin: Alright, I will propose something on the list and submit an update. This was always gonna take a while and have difficult things. So it's actually been less controversial than I expected it might be.
Colin: So this is again something we've been talking about for a while. We spoke about it just before Christmas again, most of it people seemed happy with Alvaro had some comments about phrasing and he's just using sort of defining as covered individuals terminology and using that in the same way the IAB and IESG policies do. I did that and I think it actually helps clarify the wording significantly. I clarified a bit it said if there was, sort of describing the appeals path, it said sort of appeal to the IAB by the IAB chair and I just clarified that just say appeal to the IAB. And the other thing I added was again, Alvaro had some comments, but he said there wasn't a discussion of the disclosure of procedures and he was right. I added two paragraphs at the end starting disclosure and recusal procedures which I copy pasted from the IESG policy changing IESG to IRSG and adding a sentence saying the Secretariat would maintain a public record of disclosures. But otherwise it's I think identical to the IESG policy. So those two paragraphs I think are the main ones that need review. The rest, I think people are happy with, at least based on the discussion before Christmas. So if people have had a chance to read it, I know it was fairly close to this conference, then I'd appreciate feedback and if not on the list and I see there's a couple of minor nits in the document so thank you for that.
Mirja: Can you maybe share the link again in the chat?
Colin: I think that there's a bit at the end that says to discuss which I think we've already discussed, so I'll drop that out. I think that's left over, it's the text before that that we're looking at.
Mirja: So we are maintaining this for all RG Chairs as well, right? Because I don't think we actually have a conflict of interest policy for working group chairs.
Colin: Right. That’s true, yes. It’s placed to the IRTF chair, the IRSG, so that includes the research group chairs and the ANRW TPC chairs and organizers of any other workshops we have.
Mirja: Yeah, maybe we can actually at this point, think about adding tooling to the Datatracker to record the disclosures in there.
Colin: That's possibly something we should add to the IAB and the IESG ones as well. We should try and get a common mechanism for all three bodies. Otherwise we’re just going to get in a mess.
Mirja: Especially if people are in multiple bodies and then you have to close the same thing multiple times.
Colin: I mean for the IRTF chair there’s the IAB policy as well. I agree getting that into the Datatracker might be useful. I think I suspect for now it will be a page on the IRTF website, same as it on the IAB and the IESG sites, but tooling would be a good idea. I'm not hearing strong objections that could just be because no one's had time to read it, so please do read this what what I would guess will happen as if there are no great objections is that maybe in the next meeting or just after Bangkok, we will sort of run this process, get the disclosures and sort of bring the policy into first and we can, I guess Dirk and I will talk about what when when when we adopt this or if Dirk is gonna turn around and say no this is all nonsense as soon as he takes over.
Dirk: I think this is good to have for sure.
Colin: If we do it in the next meeting I have to disclose for a fortnight's worth of the effort. I'll just point to the IAB disclosure, this is easy. Alright, so please do review this and send comments over the next couple of weeks. I think we are now in reasonable shape here. I think we've addressed all the main comments.
Jenny: The last thing I had was, is not a discussion topic, it's just a reminder that session request closes this Friday. I will also be sending another reminder through email for everyone to see which groups has not put in their session request.
Colin: A couple of other things I had very quickly the ANRP talks about so we announced all the ANRP winners earlier this month.
The talks for Bangkok will be Changjie Wang talking about LLMs for Network Configuration, and Henry Birge-Lee, and I need to learn how to pronounce these before the meeting, who'll be talking about controlling BGP Path Selection via Coordinated Route announcements and the papers for those are on the website already and I think we're getting close to having a schedule for the rest of the talks as well, but we've certainly got the talk scheduled for Bangkok.
Colin: We've also been sort of Dirk, I and Jenny have been chatting about opening the Diversity Travel Grant call for the Madrid meeting early and to help give people time to get visas. The current plan is to open that call for travel grants on Friday.This week gives three or four weeks for people to apply so we can make the decisions before the Bangkok meeting and give people a bit longer to organize visas. And then I think the plan was to open the call for travel grants for the Montreal meeting in November.Basically straight after Bangkok. So we're running sort of one ahead. So, yeah, after the summer meeting, so after the spring that the March meeting opened the call for travel grants for the November meeting and just run one ahead to give a lot longer to arrange visas cause we've had a bunch of people who've had trouble getting visas in time for meetings. So if you see a call for travel grants going out this week then that's what's happening there. We're just trying to get ahead of ourselves.
Mirja: That's a really good idea. I mean it depends on how many requests you usually get, but do you have to kind of have a small or like a short period and then close them and then make a decision or can you just like decide kind of on the run and leave it open for longer.
Colin: We usually get more requests than we have money. So if we left it open for longer, it just ended up being a first come first served.
Mirja: And the people you reject are people who don’t have a good reason to be or you actually reject people where it would be nice to support them?
Colin: It’s a bit of both. Sometimes it's people who are trying it on and it's a diversity travel grant and they’re white male professors from North America and sort of wondering why they don’t meet the diversity goals. Sometimes it's people who don't have experience, we've had undergraduate applying who clearly didn't have the experience to sort of do useful things at the meeting or we've had people from industry applying sometimes. This should go to students and relevant people from Microsoft, e.g.. Sometimes yes, we have more good people than we have money for and we have to make a difficult decision.
Mirja: Because that would maybe be something to really try to invest in a little bit more and try to get a little more money out maybe.
Dirk: We're working on it, so trying to increase the budget and to hopefully get more people in.
Colin: We do ok it's not, it's not crazy oversubscribed, it's a little bit over subscribed currently so but yeah it would be good to grow things. That's all, I think all I had. The last thing I wanted to note is it's not clear that I will make it to Bangkok. I'm hoping to, so hopefully I will see you all there. I may have to let one of you organize the dinner since it's not clear I will be there, but hopefully I will make it trying to find some funding, but we'll see how things go.