ROMAN DANYLIW: Good evening, everyone. Welcome to the IETF 126 Plenary meeting here in Vienna. My name is Roman Danyliw, and I am Chair of the IETF and I'm going to facilitate us through the evening. For those of you that are joining us online, welcome as well.
As I look out to the crowd, I think just how different it really is. I think looking out, there are more people sitting here in this plenary room, which is a fraction of actual participants, and I think this is more than the number of people we had total across the entire meeting when we were here in Vienna several years ago coming out of the pandemic. So, it's really wonderful to see us back to this fabulous city and see us back in force working on the latest in internet standardization and in research supporting those activities.
This is of course an IETF meeting, and with every IETF meeting, what's applicable are the various policies and procedures to our organization, which are enumerated in the Note Well. Please take a moment to familiarize yourself with that by clicking on the link associated with that QR code and reviewing those policies if you're not already familiar with them.
We have our usual standing agenda. We will have a welcome from our host, followed by reports from our different leadership bodies, followed by the IRTF, the Nominations Committee, the Intellectual Property Management Corporation, then the Administration LLC. Then you'll have an opportunity as the community to ask questions of the different leadership bodies. The order is the IAB followed by the IESG, followed by the IETF LLC board.
This meeting, I hope you've been finding it as successful and as productive as I have, is only due to the hard work of lots of different people and lots of different organizations supporting the IETF. First, I wanted to recognize the professional staff that makes this meeting possible. That's the Secretariat, the Meetecho, the NOC, the RPC, IANA, and the LLC staff. But I also want to recognize the volunteers that run the programs inside the meeting. That's those helping with the guides and the new participant program and of course the ombudsteam that's ever present in our meetings and also is virtual. Let's give them a round of applause for making this meeting happen for us.
(Applause)
The other big thank you I want to have is to our meeting host. I would invite to the stage [Cullen] Fluffy Jennings, the Cisco Fellow and CTO, for a few words as our meeting host from Cisco.
(Applause)
CULLEN JENNINGS: Thank you. It's a great night here. I've got my popcorn. I'm at the plenary. It just doesn't really get much better than that. It will be an excellent plenary. I'm looking forward to it. So, what I really wanted to say is thank you to everyone for coming. I know that it's a huge time commitment for us to all travel and come here. I know it's an incredible amount of volunteer work that's been done by everyone to make this happen, everyone from the people who have set it up to all of our chairs, document authors, IESG, IAB, the whole group. So thank you for this.
It's really important what we do here. Cisco finds it very important to us. That's why we continue to always sponsor it. But it's really important to the world. And I think sometimes we don't step away and think about the overall impact that the internet is having across it and how much we drive that. We, being the people in this community in this room.
So, I thank you for coming. I thank you for engaging with everyone and having a great time, throwing a little popcorn, maybe. And have a great night.
(Applause)
ROMAN DANYLIW: On behalf of the IETF community, we're grateful for Cisco's support. So I'd like to present this plaque to you on behalf of the community, to you and the company.
CULLEN JENNINGS: Thank you.
(Applause)
ROMAN DANYLIW: Very good. Next up on the agenda, I wanted to go through the chair report and the IESG report. There's a few details that I'm going to surface that I'd like to make sure that everyone sees here tonight. This is of course only a subset of what has actually been summarized and reported. For additional details about everything I'm going to talk about and more topics, especially trends we typically report that are not here, please do check the standing report at the link there. It will also be linked in the proceedings with significantly more information, posted in advance of this meeting.
You probably heard me when I opened with a little bit of excitement, this is a huge meeting. The turnout is about 1,900 people, which I had Cindy from the Secretariat look up what that means in terms of the historical baseline. The last time we had a meeting that had 1,900 people, it was San Diego, December 2000. So, quite a long time ago.
(Applause)
Maybe I should ask, who was in San Diego at that time? Nice. Yeah, my first IETF meeting was the one in Pittsburgh right after that. So, exciting. So by all the numbers, huge turnout here. Like I said, amazing to see you, given last time we were in Vienna, equally as tremendously large hackathon by participant count. I think if you think about 750 registrations for hackathon, we've had much smaller meetings than just the hackathon participation alone. So it's wonderful getting that running code and a huge number of projects also coming in.
We're seeing consistent trend lines that if you look around to your right and left, there's a cluster of five of you. Statistically, one of you has never been to the IETF meeting before. This is roughly what we always see. For those of you that this is your first meeting, welcome to the IETF. We can't wait to see you at the next meeting as well. Hope to see you in some working group.
I've already said, the last time we were even this close was almost 26 years ago. So we have blown by the high water marks we typically track in recent memory, where it's incremental growth. So, huge turnout here. Very, very exciting.
To report on the trends of where everyone is coming from, we're seeing broad participation from lots of places in the world. One thing I wanted to dive into this time because we have a unique opportunity that the administration, the IETF LLC just published the community survey that I hope many of you responded to, that gives us the broader contextualization of who is the community. Jay is going to talk in his report more about that. But two things I wanted to highlight is while we have a huge turnout here today, that is but a fraction of the overall IETF community. Our mailing list and the engagements we have beyond the meetings and the engagement in the interim meetings, we're pushing about 50,000 unique identities. We measure them by email addresses, so there's probably a little bit of duplication. But we're talking about tens of thousands of people that make the standards and research possible in the IETF and IRTF.
The other observation we have from that community survey is that about half of them continue to participate in that way through our interim meetings and through the mailing list and do not come to the plenary meeting as reported. They've never shown up at any plenary meeting at all and are still participating, which is phenomenal.
The other trend that we see repeated I want to talk about is we talked about raw size, where they're coming from, let's talk about the stakeholder community. The charm of the IETF is that we convene a multi-stakeholder forum. Our community survey really repeats what we've observed over many years, that about half of that participation is industry, about a fifth is academia, and then government and civil society is about 10%, respectively. If you're wondering what that gray means, what does “others” stand for, well, what are these other stakeholder communities? Other means there's a very, very small fraction of individuals that said I have retired and I don't consider myself having a stakeholder community. There's also a large number of individuals that said “prefer not to say.” So, that's the 7% outlier there.
Moving into IESG business. There have not been any new statements issued since we last convened at IETF 125. There's been a number of announcements. I'm going to talk about the specifics of each one of them. So I won't dive into it. But the one I did want to showcase is the tagline you see at the bottom. The NomCom is going to be speaking to you, I think, in two talks to pick the future leadership. We'll have a Q&A with leadership across the table. Very shortly in the fall, we'll be starting the process to replace many of those faces that you'll be seeing sitting at these tables. Your input for that is crucial. If you're interested in potentially serving in one of those positions, being an area director, the IESG will be having an ask me anything. Come get that personal perspective. Have one-on-one conversations about what that job is like, see whether that's a good fit. Maybe that will convince you to stand. So join us tomorrow for that.
To report out on appeals since the last time we got together, there were six new appeals filed, five related to activities in the TLS working group, and one filed around an activity coming out of the IRTF.
I have this podium to talk about many things, but I also run the general area. I always like to showcase what's happening there, primarily because that affects each and every one of you. It's cross cutting across all working groups. So, I would always invite you, if you are wondering how the process is related to assigning code points or you have ideas about how those processes should change, please come to the IANABIS working group, which is the first one listed there. If you're interested in evolving the processes related to the internet standards processes, that's what the PROCON working group is doing. PROCON is meeting tomorrow. You missed IANABIS, but you can catch up. That convened earlier this week.
To tease something that's going to be coming, the outward-facing processes by which the IETF interacts with liaisons to other standards development organizations, which is crucial to building things like digital public infrastructure or large-scale deployment on the internet, we have processes, of course, that are RFCs that guide that. They have recently been revised by the IAB. They're now coming into the IETF stream for approval. So I'll be coming to you, to the community, for that feedback to see whether we have consensus to publish those.
The big thing I want to announce that's happened since IETF 125 is that RFC 9945 got published. That has defined the new moderation process for the community. We are now in the process of implementing that guidance. The first of those things to implement is defining and appointing a moderation team. The IESG has found these [six] extremely capable individuals. We have Chris [Box], Martin [Duke], Bron [Gondwana], Susan [Hares], Murray [Kucherawy], and Mohit [Tahiliani], who have graciously volunteered to be the first moderation team, and they will be the ones responsible for defining the procedures by which RFC 9945 will be implemented, and ultimately this will go to community feedback and have to be approved by the IESG before being in operation. So, I'd really like to take a moment to recognize the tremendous request and load we have put on this crucial role and the team. So really, thank you for volunteering for that team. I'd like to recognize them.
(Applause)
And one other adjacent announcement I'd like to make is that implementing that moderation process tasks the IESG to define a forum where moderation should occur for the community, and that has been designated as the former mailing list of the ModPod working group, mod-discuss. That is the place where we should bring community conversations associated with that. As the moderators begin their deliberation with their processes, I'm sure there will be more communication to the community on those matters.
And another team that had a recent change in personnel that I'd like to highlight is the crucial Ombudsteam. The Ombudsteam focuses on ensuring the consistent application of the IETF anti-harassment policy. They are a resource to us here in the community to handle situations where you may have observed harassment in the organization or you, yourself, feel like you have been harassed. They're a resource you can come to, to work through those claims and ensure that does not happen inside the community. I'd like to recognize an outgoing member of the Ombudsteam. Sofia Celi served three years as an ombuds on the team. Equally, I'd like to recognize Padma [Pillay-Esnault] as the incoming member of the Ombudsteam. She'd just agreed to join. So, I'd like to recognize them: Sofia for her service and Padma for her willingness to serve in the community for us.
(Applause)
And last, the Ombudsteam asked me to add the bottom information. They are a resource to you, whether you need them for things that are happening remotely or whether things are happening here on site. So if you'd like to reach the Ombudsteam, you can always get them through that email alias you see there. If you want to reach out to someone remotely, Avri [Doria] and Shivan [Sahib] are standing by remotely. We also have part of the Ombudsteam here on site. So you can find Padma or David [Schinazi] if you want to talk to someone in person. That alias gets to the Ombudsteam whether they're remote or on site. But if you want to talk to a human down here at the venue, they are circulating around with a little tag on their badge.
The last thank you I would have before I turn it over to the next speaker is to really recognize the hard work of your ADs that are working the standards process on your behalf, whether moving documents through your working groups or helping you spin up new working groups. That's frankly only possible through the generous support of, largely, their employers. Not everyone is employer-sponsored, but those are represented here. So I wanted to take a moment to signal appreciation for those organizations supporting them. So, one more time.
(Applause)
With that, I'd like to turn over the stage to my colleague Dhruv, who's going to come up to talk about the IAB.
DHRUV DHODY: Hello, everyone. I'm Dhruv. I'm the IAB chair. We also have a much more detailed report that you can find on our proceedings on voteiab.org, as well as on the IETF 126 proceeding. We also had a session called IAB Open, which happened yesterday. If you're interested in IAB activities, look at the recordings for that and the proceedings. You'll find a lot more details than what I'll cover in my short report right now.
One of the activities that IAB does is organize workshops. A final step in the workshop is we publish reports. We have published our reports for three of those, and the final workshop that we did last December, even that we have shipped from IAB side, and it will be published soon.
IAB is also responsible for the liaisons. We recently established a new liaison relationship with Optical Internetworking Forum, or OIF, and we wanted to highlight that. As Roman was mentioning about some of the updates in the document, those are things still currently in the IAB stream, and we'll be passing it on to Roman pretty soon.
IAB is also responsible in tracking and coordinating outreach activities with the IESG. We undertook some of these activities since IETF 125, of which I wanted to highlight a few. We went to operator communities like RIPE, LACNIC, as well as NANOG, and we have plans to go to APNIC as well in the coming months. We also participated in some other events, some IAB members and leadership, like RSA Conference, where we talked about some of the working groups' output, as well as a free IETF event in India called Connections.
Finally, we are also responsible for appeals. Since IETF 125, we received three and responded to all three of them. With that, let me pass it on to Dirk.
DIRK KUTSCHER: Yes, hello. I'm Dirk Kutscher, the Chair of the Internet Research Task Force. Let me give you a quick update on what's going on in the IRTF.
The IRTF has 16 active research groups, and 14 of those have been meeting or are still meeting this week. I promoted the SUSTAINRG and the SPACERG research group from a proposed group to a regular group this week. We have two platforms for raising the profile of the IRTF in the academic research community. So this is to bring in new people from the research community to the IRTF, get in new ideas, and so on. So that's our workshop, ANRW, and our research prize, ANRP.
The Applied Networking Research Workshop, ANRW, took place at this IETF on Monday. So we're doing this in cooperation with ACM, supported by Akamai. The chairs this time were Thomas Schmidt and Suresh Krishnan. You can find all the papers and presentations on the IRTF website.
The Applied Networking Research Prize is to recognize recent and relevant academic publications that are relevant to, you know, the evolution of the internet. We are having like a call for contributions. We select the best papers, and then we invite those people here. This is supported by ISOC, Comcast, and NBCUniversal. This week, we have two of these prize winners here that will be presenting in IRTF Open tomorrow at 14:00 p.m. That is Rumaisa Habib for her work on formalizing dependence on web infrastructure, and Diwen Xue for his work on fingerprinting deep packet inspection devices. So that's going to be tomorrow at 2:00 p.m.
We are also offering diversity travel grants to bring in people from different backgrounds who could find it difficult to attend our meetings. So, at this meeting, we had the pleasure to host four travel grant winners from different regions and different backgrounds. We have already made the selection for IETF 127, and we will open the call for IETF 128 later in November. So if you're interested in that, please keep your eyes open.
I'm happy to say that the IRTF is recently doing quite a few interesting outreach activities. So we are going to other communities. So, for example, the SUSTAIN research group held a meeting last week in Germany that brought together internet researchers and people who are working on electric grids. So how can both communities learn from each other and also what are research interests for us but also for them.
In December this year, we are holding another workshop on inter-networking for AI at ACM CoNEXT. That's it for me. I'll hand it over to Benno and Mankamana for the NomCom report.
BENNO OVEREINDER: Yeah, you don't expect to see me here this time. Here I am. So, I'm the NomCom Chair of last year. And as it happens, there was a vacant position on the IETF LLC board. That's still the duty of the NomCom of the previous year. So here I am and briefly to report on where we are right now.
So -- there we are. It's just myself here. So, there was an IETF LLC Board position vacant. And that means that we have six weeks to announce a nomination. We have, of course, interviews, we have deliberations, the confirmation, and then the install of the new board member.
Six weeks is not long. So that was -- time goes fast. Where are we now? First, I want to thank the flexibility and availability of my NomCom. There was some muscle memory. So, things went fairly well and smooth. Thank you, all, my NomCom members. But again, where are we now? What's the status?
So we had community feedback up until Tuesday. We had interviews this week with two nominees. This morning, we had our deliberations and came to a selection. We sent the letter to the IESG with our selected candidate, and that's where we are now. We sent the letter to the IESG, the confirming body, and approval, confirmation will arrive soon. And that's it. So stay tuned, and we'll soon hear more about the next IETF LLC board member.
I have to mention, this is a situation that the vacancy is less than eight months, less than a year, I have to say. So what happens is that the old NomCom appoints or selects a candidate for the midterm, which is less than one year, and according to RFC8713, it's also a selection for the next three years for that position. Okay. Any questions, comments? Maybe later. I want to hand it over to Mankamana. Thank you.
(Applause)
MANKAMANA MISHRA: Good evening. When Roman started the meeting with such high energy, I can't match his energy. But I'll go slow. So I'm Mankamana Mishra, NomCom Chair for 2026, 2027. And what exactly NomCom is responsible for is to select candidates out of nominees for leadership positions. We have to go through all the process of getting nominations, getting the feedback, having interviews in the next IETF.
Right now we have total positions open -- thank you, Benno, for taking care of the top position, so I have one less thing to take care of. We have one position for IETF Trust. Then IAB has six positions. IESG has seven positions, one for Applications and Real Time, one for Internet, Ops is one, and Routing AD, two, Security AD, one, and Web and Internet, one position.
We have already finished having the voting members. So we have done the ten voting members before this IETF. We have non-voting members, where I am the chair, and the Past Chair, Benno. Then we have some liaisons from IAB, IESG, ISOC. There are two of them missing, which we expect to finish soon.
With respect to timeline, NomCom has already seated this IETF. We had our kick-off meeting on Tuesday. From now to IETF 127, we will start nominations from the 1st and it will be going on until 14th of October. Looking at the next IETF, some of the people may not travel, so we'll start virtual interviews pretty heavily around 20th of October, and we will have in-person interviews during next IETF. Definitely I would request everyone who is nominating themselves, if possible, please travel. It's always good to meet in person. And IETF 127, we are going to have deliberations, and we will have all the positions filled by IETF 128.
Based on the experience of last IETF, last time the Friday in Montreal, we were able to, I would say, finish almost 80% to 85% of the work. So we're hoping the same thing will continue in the next IETF as well. The detailed timeline is already -- we have put in the Datatracker wiki. Thank you.
(Applause)
WENDY SELTZER: Good evening. I don't see nearly enough popcorn in the room. Wendy Seltzer. I'm here as President of the IETF Intellectual Property Management Corporation and Chair of the IETF Trust. I am helped by this tremendous board of the IPMC. Victor and Jon were recently joined by Lars and Sean. As it's my first time up here, I have to say thank you very much to Glenn Deen and Kristin Berdan for their service previously on the IPMC Board.
IPMC manages the intellectual property, the copyrights and trademarks, for IETF and also the IANA intellectual property. Our primary duty is to safeguard that for the relevant stakeholder communities. So we report to the community on licensing inquiries and legal actions. I'm very pleased to report we have not had any legal actions or inquiries that required a detailed response.
We are still in the process of the transition from IETF Trust to the IP Management Corporation. And we have transferred all of the IETF-related assets at the previous meeting, including future-acquired assets. So, don't worry if you're still making agreements that transfer your intellectual property to the IETF Trust. It is going to its proper home with the IPMC.
The IANA IPR will be transferred after we complete final signatures on novation agreements to make sure all of that is recognized by all of the parties who have interest in that IP. We are circulating a new TLP. That was the Trust Licensing Provisions. It's now the Technical Licensing Provisions. If you're interested in the wording of that, please look at TLP-interest@ietf.org. If you have comments on that, please share those. We're also planning a BCP update and have a draft linked from the slides.
We have, as was mentioned, one seat for appointment by the NomCom. If you're interested in that, the job description will be very similar to those previously. If you have questions about that, feel free to come to our office hours tomorrow or to reach out to us through our websites and email, as below. Thank you, all, for helping to keep our intellectual property in good hands and for continuing to produce valuable work that we can help to safeguard.
(Applause)
JAY DALEY: Hello, everyone. Welcome to IETF 126 Vienna from the IETF LLC. I hope you're enjoying this venue. It's a smaller venue than most, but I think many of us are finding that's lovely because we don't have to walk long, huge corridors to find people. And certainly getting to meet people a lot better.
Talking first about IETF 126 Vienna. Roman and I will at some point finally get together about these things, but for now, if I could ask you, please put your hands together and thank Cisco for being our global host.
(Applause)
And also, if we could thank our gold sponsors for diversity and inclusion. We have Akamai and Huawei.
(Applause)
And then our many other sponsors. We have Comcast and ICANN, DENIC, Nominet, ZTE, and connectivity sponsor and local supporter is Next Layer. Thank you all very much.
(Applause)
These are the people who actually run the meeting. If you have any thought that I didn't know, these are the people that did it all. So, the Secretariat, thank you all very much.
(Applause)
And these are all the supporting technical teams, both the staff and the contractors and the volunteers. We've particularly called out the companies that those volunteers are from who support them coming here. Thank you.
(Applause)
And then the final clapping slide. Thank you all very much to our global host. It is with their ongoing support that we're able to afford these meetings and have such delights as popcorn. Thank you so much.
(Applause)
Some new things for this meeting. First of all, you may have noticed the Secretariat are no longer wearing uniforms. Instead, we have now the staff blue badges around so you can recognize and identify people a little bit better. We have a new emergency plan in place as well, which is for a range of emergency situations, and everybody have been trained on it. Strange enough, we had to use it on Friday when we had a very large thief stealing NOC and Secretariat consumables. Stealing something of the NOC is really a silly idea, so that didn't last too long. By the way, they were hotel guests and had to be removed and everything. So, good job we had the plan in place for that one.
Once again, just to remind you we have a store. T-shirts and other things are available for you on the store for you to buy if you've missed the chance this time or need extra sizes or anything like that.
A little bit now about IETF LLC operations, just to update you. We've had some staffing changes and things with the Secretariat being insourced. The Secretariat has now been restructured slightly into two specialist teams, the meetings team and the leadership support team. So when you put the organization together, this is the rough structure now of the organization.
We have a new data analyst as well working on multiple projects and our community survey that just came out recently. And we now are getting much more strategic with annual planning sessions for each team with a follow-up community report, a blog post generally, for you to understand more detail about what's going on with all of those teams and their plans.
We've got some upcoming consultations. We're going to be doing one or two, actually, about backlogs and SLAs. So both the RPC and Tools teams have significant backlogs of community work to process. So we're going to do community consultations on your expectations of service. What's done, what performance is expected. And if we find there's a mismatch between community expectations and the resources we have, then that may lead to conversations about that.
We're also going to have a consultation on the remote rooms experiment. So some of you remember that we did an experiment in IETF 125 Shenzhen about remote rooms. No remote rooms were requested for this meeting, but we are going to do consultation on the whole concept before IETF 127 San Francisco, because we suspect if it is then available, some people will be interested in that.
We're also going to be putting out some venue assessment reports and consultations on new venues, new cities on IETF meetings coming up. Overall on that venue side of things, we'll be working much more on transparency and engagement with the community about how we choose venues and things going forwards.
This is the key findings from the IETF community survey that Roman mentioned earlier. The first one is a really very, very strong one. First of all, I want to mention we have roughly 52,000 email addresses subscribed to IETF mailing lists. We send this survey to every one of those addresses, except maybe -- well, several hundred have opted out of it from previous surveys. We had 1,480 valid responses after we do our data cleaning and stuff afterwards. One of the first things we do is ask people to self-classify how often they engage with the mailing list, and that gives us a measure of how much of a participant or how regular a participant they are.
The IETF continues to deliver on its core mission with perceived importance at an all-time high. When we ask people how well do you think the IETF does, is it doing what it says it does, very, very good results all around there.
Then we have one of the bigger concerns, which is about the timeliness and process right the way through. So that's the decision making within working groups, that is the publication side, it's all of these things. It's a persistent unresolved concern there.
We are making meaningful signs of progress on gender diversity. This is particularly because of the new intake, the younger people that are coming in. We have a much more diverse set of people coming in than the long-term participants.
We understand the process complexity and reading volume are a hindrance for new participants. That's been well understood. We have the whole day Sunday training program, and we're looking at more initiatives as well to help people come in and become a productive member of the IETF as soon as possible.
Then we have the number six one. We still have people feeling that balancing this corporate extensive representation with the core principle of individual contribution is a challenge. It's a gray area for many people about how that sits.
Then finally, we, for the very first time, had an open-ended question. What else do you think, on the end of the survey. We tried to avoid it before because I didn't want to read 1,480 responses. Then along came Claude, and we got Claude to read 1,480 responses for us. We only actually had 240 responses. So it was much easier than expected. And amongst -- within that, within a smaller group within that of regular participants, we had some governance and inclusion concerns surfaced. So, it's a very detailed report. It's very well worth reading. It's very useful.
If you have further questions from it, please let us know because we're always happy to try to dig into the data more. And we also can make that data available to researchers if we, you know, can do a proper research agreement around it. So we have that there as well.
Right. So a little bit about future meetings now. Next, of course, we have IETF 127 San Francisco. Registration is open. Please register now. Please book your flights. Please, let's see you in person. That would be lovely.
Then if you haven't seen it, we have now announced IETF 128 next March will be in Kuala Lumpur. It will be in the Kuala Lumpur convention center.
(Applause)
Thank you. One person said to me, yay, proper Asia, which is great. So, very pleased with that one.
These are then the future meeting locations and venues. The next four are all booked. We have San Francisco, then we have Kuala Lumpur, then we have Berlin. After that, Vancouver. And we're now starting to work on the next set going forward. It did take a long time to book Kuala Lumpur. Sorry about that. It was quite late. Getting new convention centers is always very difficult. It takes -- well, it takes a very long time of backwards and forwards around contracts and liability and things. Every convention center we've ever come across wants us to take liability for absolutely everything and pay them extraordinary amounts in advance and these things. It's always a painful process to go through. I'm not quite sure how other people work with convention centers, but that's the way forward.
So that's it. If you have any thoughts or views about locations, venues, this venue, how it's working for you, please let us know. We will have the survey at the end of the meeting. As many of you as possible, please fill that out. That would be great. We're particularly interested in the size of the venue. This is a smaller venue. We were worried about the capacity. The capacity is 1,350. We have got 1,250 roughly registered, maybe 1,100 on site at any one time. But for many people, the feedback I've heard from most people is that this is working much better than a larger venue and maybe we ought to try more regularly to find smaller venues and do this.
So, all right. That's just the state of things forward. Thank you very much for your time, everyone. Over to Mirjam.
MIRJAM KUHNE: There we go. Yes, my name is Mirjam Kuhne. I'm the Chair of the LLC Board. We have one other board member here, that's Roman. Unfortunately, George and Gonzalo couldn't be here this time. And as you heard from Benno, we have one vacant position, and you heard the update of that. Luckily, Roman has agreed to take over the treasurer role since our treasurer resigned. So we're happy with that. And I would really like to thank Shauna Turner for the work she's put in on the board since she started.
As usual, just an overview here of our financial situation. This is an update from April. I know we're always a little bit -- not behind, but this is how the statements work. You can see the monthly financial statements on the website. So the May statement is already up there. And the board will be approving the next ones coming up. I'm not going to go through the details here.
Also, on the next slide, you'll see some notes, some annotations there for the numbers, if you want to go through them. Please feel free. If you have any questions, we're here, Jay and also the financial officers.
Then I would like to give you a quick update of what the board is working on, in addition to all the operational work that Jay has described. We had a retreat, a board retreat, that we do usually once a year. This one was in Amsterdam in April. A lot of work there and I published a blog post about this after the retreat. We're busy following up with all the action items and all the decisions and objectives we want to work on. We already have agreed on a new or have reviewed our risk register and accompanying risk policy that describes the risk register. That's already updated and on the website.
We are also updating our investment policy, which is a healthy thing to do every so often. So we're working on that. Also, our multi-year strategy is up for review. So we're working on that at the moment. Jay has already mentioned the remote rooms. So we've been talking about this also on the board. We carefully read the report of the last one. So we're looking forward to the consultation there and what you -- how you would like to go forward with that.
And I started -- and I heard people want to find out a bit more of what the LLC is doing and also what's happening on the board, what the board members are thinking about, what decisions are made. So we started publishing a report just before this meeting. I sent out the first one and would like to continue to do this maybe on a monthly basis, maybe sometimes in the summer it won't work. But I'd be curious to hear if that's something you find useful. And pulling together the other reports that Jay and the other teams are sending out, like the IRPC and the Tools team and some other activities going on within the LLC that maybe they're not all published in the same place. So I thought it might be useful to do that and send that out, a bit similar to the IAB and IESG reports. We want to do that also for the LLC.
We also had some office hours yesterday -- Monday and Tuesday. Roman and I were there just to see if there's any feedback, any questions you might have for the board. We might repeat this from time to time, see if there's interest in any topics you want to bring up with more time, rather than here in the plenary.
And that's pretty much all I had to say. Here are, as usual, the times of our board meetings. Also, parts of it are open to the public. We usually have a number of observers there. We have the dates. We're going to skip the one in August, but we'll be back in September and October. And you'll always find the minutes of the meetings on the website. So you're welcome to join as an observer.
Then on the last slide, just a bunch of links and how you can contact us and reach us and different teams -- board only, board with Jay -- and also links to all the information that I mentioned, all the policies, financial statements, minutes, and so forth. And that's it from me. Thank you.
(Applause)
ROMAN DANYLIW: Next up, we're going to start the community feedback with open mics from the leadership body. First up, we're going to have the IAB. So if you're on the IAB, please come to the stage.
Yingzhen Qu, IAB.
Dirk Kutscher, IRTF Chair.
Yaroslav Rosomakho, IAB.
Roman Danyliw, IETF Chair.
Matthew Bocci, IAB.
Suresh Krishnan, IAB.
Dhruv Dhody, IAB Chair.
Alvaro Retana, IAB.
Warren Kumari, IAB.
Jason Livingood, IAB.
Ali Begen, IAB.
Nick Sullivan, IAB.
Jana Iyengar, IAB.
DHRUV DHODY: And the mic lines are open. Okay. We have one.
ELIOT LEAR: Good evening. This is Elliot. I can't let you guys get away without at least one comment at the mic. I am the Independent Submissions Editor. I just want to bring to the community's attention that I will generally keep drafts up to date in the tracker if people want to follow what's going on with independent submissions. There's a group of independent submissions that are now being sent off to the IESG for 5742 review. These are the CNSA2 crypto profile documents. You can find them in the tracker by searching on CNSA2. I solicit comments on these and any other drafts, and it's not too late to comment on these. I'm always looking for reviews, and please feel free to send me any comments or questions you might have. Thank you.
DHRUV DHODY: Thank you, Elliot. Thanks for your service. Any other questions for the IAB? Going once. Going twice. Yay.
(Applause)
ROMAN DANYLIW: Next up, I invite the IESG. Gorry, why don't you get started with introductions.
Gorry Fairhurst, WIT AD.
Chris Inacio, Security AD.
Charles Eckel, ART AD.
Eric Vyncke, INT AD.
Mahesh Jethanandani, Ops AD.
Roman Danyliw, GEN and Chair.
Dhruv Dhody, IAB Chair.
Gunter van de Velde, Routing AD.
Ketan Talaulikar, Routing AD.
Mohamed Boucadir, Ops AD.
Jim Guichard, Routing AD.
Deb Cooley, Sec AD. Isn't that routing?
(Laughter)
Tommy Jensen, INT AD.
Mike Bishop, Web AD.
Andy Newton, ARCT AD.
ROMAN DANYLIW: Okay. The mic line is open. I see some queued up. So, Wes, I see, is first.
WES HARDAKER: Thank you. I'm Wes. I've been coming here for a while to these IETF things. One thing I've noticed, that there's rather large waves of pleasant times and times with a little bit of discontent, where we -- I don't know if you've noticed, we don't always get along. There are times where in disagreement with each other, there are times where we're in disagreement with the chairs. There's times that we disagree with the ADs. So, I come here with a simple message. First off, I hate all of your “discusses”.
(Laughter)
But the reality is that we all come here to make the internet a better place. What we often forget is that we each have a different vision of what better looks like. And so, we have different opinions, we have different priorities, and we get in fights, but we're all here to make the internet better, more resilient, more secure, more robust. Use all the fancy buzzwords you want to use. We all have that goal.
My message tonight is simple. That when we get in large disagreements and when we start fighting with each other and when we start picking on people in unfair ways, it's not fair. So I want to say thank you. Thank you for your service. Thanks for putting up with us. Thanks for the time and devotion that you have. Thanks for making the internet better. And thank you even for the discusses you put on my document because I know you're trying to make the internet a better place.
And I would say I'm sorry, but I hopefully don't have to. Other people should. So, thank you for all of the people that don't say that they're sorry. Thank you.
(Applause)
MIKE BISHOP: I don't think any of us enjoy putting “discusses” on documents.
(Laughter)
MIKE BISHOP: Okay. Some past IETF chairs may, but...
ROMAN DANYLIW: Next up, Martin.
MARTIN THOMSON: Thank you. I have three questions. I don't know how you want to receive these things, but really, this is a request for you. The first one relates to the matter that's been happening in TLS. I think there's been some truly disgraceful behavior involved in that process, and the chairs have done, I think, the best they can with a terrible situation. I thank the chairs for that work.
I think two things are true at the other end of this. I personally like the outcome that the chairs have ruled, but I also don't think that this is consensus. And I realize that we don't have a process for dealing with situations like this very well. And I think it is the responsibility of the IESG to deal with the situation and give working group chairs better tools to deal with these terrible situations where we have this sort of essentially denial of service attack on the organization. So that's one.
The second one is the flood of -- I'm going to be frank about this -- AI slop that we're seeing in particularly dispatch and a larger number of working groups. This institution has for a very long time relied on certain defensive measures to protect itself from having to deal with work that consumes time and is ultimately not productive. I would like to ask the IESG to do something more substantive about that because those defenses that we had are no longer adequate to the task.
And the third one is a bit of a criticism. Somewhat related to that last one, I see we have, I think, approximately four AI-related BoFs this week, three of which have already run, I think somewhat unsuccessfully. I feel that the IESG scheduled those BoFs far too early. The work was not mature enough, and I would like to see a little higher bar used to filter work like that so that we don't spend enormous amounts of the community's time on work that is ultimately ill formed and not quite ready for the community at large. Thank you.
ROMAN DANYLIW: Thanks for those three comments. I'm going to start responding to the three backwards. So on the observation on the IETF -- on the AI-related BoFs that happened, the IESG, with help of the IAB, certainly deliberated on those topics, and we can never predict what the future outcome is of those BoFs, but in each of those cases, we felt that there is at least sufficient community conversation to take it out of the forums in which it was happening, inside meetings and other places, and give it a structured place to discuss with the entire community where we can get the consensus feedback, whether these kinds of conversations can continue.
Anyone else want to add on to number three? Okay.
MIKE BISHOP: Can I throw one thing in?
ROMAN DANYLIW: Of course.
MIKE BISHOP: I think Tommy wanted to as well. I will also note that a successful BoF is not necessarily one that decides to form a working group and do work. A successful BoF -- it is still a successful outcome to decide we have more questions that we need to answer before we move forward. That's not necessarily a failure.
TOMMY JENSEN: And on that note, this was definitely -- these were not easy decisions, and it was something we were attempting to balance in terms of if we -- what will happen if we give a better venue for discussion because there was a lot of criticism and a ton of side meetings going on that indicated there was something that needed to be broken out. And if the outcome is that it all comes out and turns out no actionable work is there, well, now we know. Yeah.
ROMAN DANYLIW: Okay. Then addressing the middle one, which is of course the easiest of them all. AI and the slop coming. Listen, we got to acknowledge outright that AI is already changing how we do the standards process, and it's going to change it in ways we can't even imagine. It's going to change how the drafts are written, how the drafts are reviewed. It's changing already how drafts are -- I shouldn't say future drafts. It's happening now, and it's going to happen in the future. It changes the whole dynamic on running code when I can mint a lot of code real quick. It'll change how we interact and is changing how we're interacting on mailing lists and how the meetings run. Then lastly, it's also impacting the consensus process.
The IESG is hearing from, at the participant level, at the directorate review level, and from working group chairs that what we have is inadequate and we need to sort some of these things out. Equally, the most recent version was the working group chairs' lunch. We heard practical -- I loathe to say best common practices -- but practices around how we can exploit AI in the standards processes. We need to be considering both. The IESG realizes we have probably not facilitated that conversation enough, and in some of our prep meetings coming into this, we recognized that.
In the coming months, we're going to scrum to figure out how we can create a forum. The easy answer is make a mailing list, but we feel we need to do more than that. We need to create some facilitated structure to get community input and community decision making to come up with either the practices, the norms.
We don't really understand what the shape of that is going to be, but we want to make sure we have a forum, we bring the community in, and we want to think a little bit about that. If folks have ideas on how to convene such a forum. And we're not sure what comes out of that. What does the practice look like? Is it an RFC? Is it more? Is it less? What exactly is it? But we realize we need to have a structure to think about both how to harness AI, accelerate where those are, codify that, make that broadly aware, and equally, it's impacting negatively the standards process, so how do we intervene as a result of that.
ROMAN DANYLIW: And then the first one, I wrote down better tools. I think it was for appeal processing, moderation. Do I have that right? Better tools for TLS.
MARTIN THOMSON: The TLS process was a disaster. And obviously, the chairs did the best with the tools they had. But they were obviously -- those tools were inadequate to the task.
ROMAN DANYLIW: I agree. I think certainly one of the things we're going to get, and hopefully, there are many dimensions to the TLS I wouldn't begin to unpack. The one I think I have an easy answer for is there are observations around what's happening around Code of Conduct. When we will have the new procedures created by the new Moderator team that comes from RFC 9945, I believe, minimally, there will be new tools there. General Dispatch, we saw an inkling of maybe we need to be talking more about how we do assessment on consensus, so it's more broadly appreciated, so we have repeatable practices. I can also see that would be a direction in future.
I don't know, I think there's one particular silver bullet approach for what's happening in TLS but given that we got five appeals from just the last time we met coming from that working group and got a few more before that, I think your IESG definitely appreciates, you know, that help is needed.
Anyone else want to add to that?
DEB COOLEY: So I will agree, there's work to be done. And the chairs and I and the rest of the IESG will work on this. It's definitely -- you're definitely correct. It is definitely out of control, and it definitely needs to be changed.
ROMAN DANYLIW: Okay. We'll take the next speaker. Arnaud.
ARNAUD TADDEI: Yes, hi, with no affiliation. Some people will be happy to know this is probably my last IETF meeting.
A few things here. First of all, I completely sympathize with the situation that you are facing on the TLS group. I have no words. I have my full support and help for what Deb -- I just could not imagine this would happen, but there is one thing regarding the Code of Conduct is that as the SG17 Chair, I had to revoke my first member of the member state. Person that, by the way, is causing you problems here as well, but I succeeded.
So, my point is that if you succeed to change that, you need to have a better life, and I can only encourage to you go in this direction because we -- all of us, basically, celebrated that this person left. That's -- it was something that happened. I can just encourage you on that.
The second point on the tooling and is there's tooling for us to help creating or modifying and so on, I must say it's unbelievable what you can do with it. So on my side, I'm encouraging everybody that we do a better job of sharing between ourselves because that's just something that needs to happen. We cannot stop it. Can we graduate it? I don't know. Can we make it under, you know, good practices? I don't know. But we are exploring. Let's share our best practice there.
Finally, on the agentic AI sessions, I am exactly not surprised on the opposite point as Martin. I appreciate that the IESG made and proposed a number of BoFs on agentic AI. I think it is necessary. We have raced against something that is taking us too fast. And I'm really surprised with some of the discussions. I'm less surprised to see some of the outcome. I would really have liked to dome to be accepted. In fact, I'm on the opposite side. I think we don't create the IESG to be more prescriptive. I'm looking at Eric here, if possible. Because you are not alone. I see two things here that do not work for me. One is the obfuscation of certain things on their stuff because they're doing that for 30 or 20 years with no way to get even a discussion about that.
The second thing is I see an inward-looking of this community completely ignoring what's happening outside of IETF. I'm glad to see Peter is here for 3GPP. I'm here for SG 17. And some of you actually recognize that. But it's going so fast. I mean, you can't just stay there. Some of you are going to leave here because they are not finding the place to develop their work, and they need it.
I can only encourage you to be more perspective accelerate because the clock is ticking. And I will stop at that. Thank you.
ROMAN DANYLIW: Thanks for the reminder on the timer. And just to be concrete on the idea of sharing best practices on AI in standards organizations, that's an excellent idea. For folks interested in that topic, there will be the recording from the working group chair session where a particular implementation of how we use AI to help in the standards process was introduced. And I think at the RASPRG, which I think is happening later this week, someone will correct me, I believe there's a couple presentations on that. So that material will be available.
Excellent. Next up, John. Remotely.
JOHN KLENSIN: Two things. I'll try to be brief. The first is that I may have missed it because I was unfortunately called away with an emergency at what may be a critical time, but I believe that we should be getting at least a report of the number of inquiries the Ombudsteam team is getting in our meeting gap, and I hope for the same thing with the Moderation team that comes around and actually starts dealing with questions. It's important information for the community to have and how much the activity level is there.
Second, and probably more important and immediate is not a complaint about any particular set of discussed votes or IESG comments. But we've had a micro-controversy since the time RFC 2119 was adopted as to whether the use of the term "should" or "should not" in a document needs to come along with an explanation what the exception cases are. And I have recently been on the side of those who feel like there should be such an explanation. But it has never been a requirement.
And I'm concerned if it's not a requirement when the IESG comes along and lists a number of places where the document doesn't do that. So this is a request the IESG clarify what they think is the situation there. And if necessary, initiate an update to be BCP 14, which clarifies what the rules are about the use of should and not should so it's consistent across the board and working groups don't get to use the term Wes might object to, ambushed by these questions or comments or complaints. And again, not a complaint on any particular issue, but we've gotten ourselves into a situation where there's a lack of clarity there, and we would be much better off having that clarity at the time working groups are producing documents rather than showing up in IESG voting. Thank you.
ROMAN DANYLIW: Thanks, John. I want to make sure I understood. I believe your first question was around you want more transparency on the processing of cases by the only Ombudsman’s team and the Mod team. And assuming that's the case, the ombuds numbers are published in the -- in my Chair Report at every meeting. So if you check this meeting's report, you'll find the numbers in there. The last meeting, so it started two meetings ago, so you'll find them going back at least two meetings.
JOHN KLENSIN: I didn't notice it there. If it's there, I apologize.
ROMAN DANYLIW: No worries. The Ombuds team started rolling those out, so those numbers are now available.
On the idea of modifying the BCP on keyword draft, the IESG is not working on personally minting a draft. We would welcome, certainly, someone from the community giving us a start point and figuring out, path finding where we should go.
ANDY NEWTON: Roman, I want to also add that the IESG has a statement on BCP 14 language.
ROMAN DANYLIW: Yes, if you are not aware, that statement, certainly, there is also that.
JOHN KLENSIN: I am aware of the statement but expresses a preference as I read it. A preference, not a rule. And again, if we're going to have large numbers of comments on a draft at Last Call time, we really, in my personal and not very humble opinion, need rule and clarity about that. Not an IESG statement alone.
MIKE BISHOP: John, can I also point out that a comment is not blocking. It's a suggestion from the AD, but you don't have to implement it. So we can comment on that.
JOHN KLENSIN: With regard, there's a funny issue, which is that working proofs are generally expected to respond to those comments rather than ignoring them. Though they may not be binding. And for one of the should case, it creates a situation in which choice number one is to modify the documents putting that explanation, and choice number two is to work up that explanation so it can be used as a response to the IESG member in order to move the document forward.
So again, we're in a little bit more -- we're in a slightly more complicated situation here than I think is generally understood. And I think it would be worthwhile to clarify it.
Again, no intention about a particular document. I'm happy people are doing their job. But this is a loose end. And the fact it's been a loose end for 30 years doesn't help very much.
ÉRIC VYNCKE: So, John, I think I still owe you a reply to your email you sent me one week or two, but you can imagine with this IETF meetings.
Regarding an update to BCP 14, honestly, I will be the first one to applaud it because it forces us, as an IESG evaluation, to be sure that RFC 2119 is really enforced. I agree, it's a guidance. And I have sometimes worked with the authors in the working group about the should or should not and the equivalent because it's quite often, most of the time, to find very easy a crystal clear and crisp specification about what they should. You can quite often replace them, but most because people are afraid. And what you can do, you can use another substitute like this. I see some people laughing. It takes time, but you can do it. And I'm pretty sure we will manage it on the specific draft, John.
JOHN KLENSIN: And again, on a specific draft, I'm not commenting. I agree with Wes, I'm glad all of you are doing your jobs. And I understand everybody is stretched too thin, including both you and me.
But as to the rest of your comment, we're in complete agreement. The only thing I'm raising now is this stuff ought to be sufficiently clarified that we don't have to iterate on it in these kinds of ways and authors and working groups are not surprised when these come across as complaints at Last Call rather than very clear guidance much earlier. So, I think we're in complete agreement.
ÉRIC VYNCKE: I think we agree.
ROMAN DANYLIW: Very good. Thanks, John. Rich, you're up next.
RICH SALZ: Hi. Of the six appeals that came through this past year, I think four had the non-conformant language that the IESG put out a statement saying is disruptive. And then they all had, oh, be in the interest of having an appeal channel, we're going to ignore it or let it through.
As a member of the community, please stop doing that. The person who does these is not learning the lesson that you have to behave according to the civil standards of the IETF. And so, the next time, just send it back. Please.
(Applause)
ROMAN DANYLIW: We are very much sensitive to the dynamic you are describing. And for this reason, we structured the guidance that would allow two levels below us to do that. But as an open standards organization, we need a defined appeal path, and we are balancing really the equity of ensuring that there always is an aperture for an appeal with what's happening in the rest of the organization. I appreciate the tension that that is creating.
RICH SALZ: Yeah, but do you really think putting an ineffective noise paragraph that has completely annoyed the rest of the community is a valid way to keep the appeal path open?
ROMAN DANYLIW: We are -- all I can say, we are trying our best to manage the consistency of the process with -- the consistency of the process with the efficiency of what's happening. Anyone else? Okay. In that case, Tim, you're up next.
TIM CHOWN: Hi. A couple of points, comments, questions. Drawing on what Jay actually posted about the survey, I think the second thing I'm going to say is for you guys. Maybe the first one is more general.
First one is about diversity. I've not been for about five meetings. And Jay was saying from the survey, it seems that progress is being made in diversity, but I see in the IAB and the IESG there are certain visible dimensions on diversity, but there's one that's clearly not there. As only one woman, I think, on both the IAB and the IESG that we've seen today, for example. I don't know whether you want to make any comment on that and whether you feel, obviously, I'm sure here is very capable of what they're doing, but what work do you think we can do to improve that?
The second is the comment in the survey about the speed with which the IETF moves, which I assume is too slow rather than too fast. What do you think we can do differently? What can you guys as ADs do differently? What can we as working group chairs and participants do differently? Are there things that we're learning maybe early from RASPRG that we can do differently? Oh well. Yeah, what can we do differently?
I would like to think there's something we can do. We seem to quite often work just work in like three huge waves of effort each year as the draft deadlines come in for each meeting. Should we be meeting more remotely? Should we have more interims? But that's effort on the chairs, on the ADs. Yeah, what do you think? Is there something we could do to make things progress faster?
ROMAN DANYLIW: Okay, so taking the first question. As Jay noted that what we're seeing in the new participant pipeline is very different than the more durable version of that. So, we believe that some of the investments we may have, in fact, improved the diversity pushing into the pipeline.
Your observation about the mismatch of the pipeline coming in and the leadership bodies is one that, yes, one does not, even if one changes the composition of what's injected in the pipeline, it takes some time to work its way through the process.
One of the ways I know I've talked about this at this Plenary meeting was one of the stepping stones to leadership positions is having the intermediate leadership positions, and the one we were focused on, the axis we were focused on was working group chairs.
One of the very conscious decisions that was made is in the cadre, I don't have the numbers necessarily at the tip of my tongue, Plenary 122, 123, the ballpark numbers for the working group chair population is something like 275, or 300. Some people have one appointment; some have two. Each group has at least two at this point. The fraction has three. So what we have focused on is reducing the number of instances where the same individual has many different working group chair appointments, but the crucial thing to provide more experience in leadership to ladder up potentially to these other leadership bodies, because the community seems to feel there's a tight link, is do the one plus one, which is you take one experienced chair, and then you have someone that's never had that role, and this gives a larger cadre of future working group chairs but also gives you more people and a larger pool to respond to that. And I believe when we pulled those numbers last when we were very consistent in implementing that policy, when I briefed that, I think a smidge under 50% was the nature of that appointment, and we were able --
TIM CHOWN: I would encourage you to report on that in the next Plenary maybe, just to give some of the evidence there. Just to encourage us all.
ROMAN DANYLIW: Okay, very good. I'll report on that. We can certainly dig those numbers again. I swear it's either the 21 or 22 Plenary has those numbers from inflight.
And then you asked about the speed at which the IETF is operating. I think that there are a number of dimensions to that.
The one that the IESG is not going to directly talk about, but Jay talked about is there's many parts of that pipeline. There's the RPC pipeline. And there's going to be a community consultation to explore all that. There's going to be deeper exploration about what makes sense in the pipeline.
If you're asking what can the IESG do to make everything before that go faster, I think there's a lot of different ways to approach that. One of the ones that we saw that we just got dual-sided community feedback was cherry picking things out of the side meetings and things that we were seeing and trying to promote them to accelerate things to either make a decision that we shouldn't be in this business or we're not through a big community process. One example that's not at all necessarily representative of everything that's happening. Mixed feedback on that.
I don't know if I have a quick answer to how to make everything go faster inside the working groups. And I would like to ask my other ADs about what are the tactical things that we should be doing to make it better. And certainly, community feedback on that.
TIM CHOWN: The amount of times the ADs put in is much more than me as a working group chair or me when I was a participant. We can dive in rather easy. But you guys, it's constant. You got your two weekly meetings. You've got everything else. You need help with reviews, et cetera, et cetera. Thank you for doing what you do, but is there anything we can do better? Any way we can help you or you help us, I think is the question.
ROMAN DANYLIW: Other ADs want to comment?
KETAN TALAULIKAR: Yeah, so this may not be representative of entire IETF and maybe some working group that I'm involved with. One of the things which would help is really people reviewing, contributing to not just their own work but other work. Sometimes we see this thing where, you know, that's not happening. And that can slow down the progression of work. So that's one observation that I have, I would like to share.
ROMAN DANYLIW: I don't want to sort of clip this question short. Let's please answer, if you have a position, but I observe a deep queue, so we have to keep moving.
TOMMY JENSEN: Just quickly, when I was looking at this from my area, after reading those verbose feedback, what I found as I tore it apart is that I was the biggest blocker. And so one of the things, just as a public, you know, we're all in this together, right? So I'm publicly saying that's part of what I've been doing is having accelerated review on my part such that it's all there beforehand, because I was the AD a couple times where a discussion came in the day before a telechat, and that's not cool. And I'm acknowledging that. Everyone in this room is volunteering, almost always, their free time, right? So I'm with you, there's not one. It's just one bit at a time.
MIKE BISHOP: And as to diversity, I would just say as -- I was on the NomCom last year as the IESG Liaison, and NomCom can only pick from the people who are nominated and who accept the nominations. If you are a NomCom eligible, please consider serving. If you are nominated, please consider accepting the role and doing the interview, even if you don't think you'll get picked. I didn't think I'd get picked, and here I am. And please, consider it. Give NomCom options because if NomCom's choices for a given position are three white men, you're going to get a white man in the role.
TIM CHOWN: Okay, thank you.
ROMAN DANYLIW: Rohan.
ROHAN MAHY: Hi. I'm not sure that this is -- whether this is for the LLC or for you guys or both or some combination.
ROMAN DANYLIW: We'll let you know.
ROHAN MAHY: I figured. Looking forward to booking my flights to San Francisco soon. I recall a certain San Francisco IETF meeting that was right before the start of the Iraq War, and there were protests two blocks from the IETF hotel, which turned out to be mostly okay for everybody involved. It certainly wasn't really a dangerous situation.
The next IETF is scheduled 11 days after the date of a scheduled U.S. election. What is the community's mitigation plan if norms are much further eroded after the -- after that date?
ROMAN DANYLIW: I want to be respectful to the question. We'll answer that during the LLC time.
ROHAN MAHY: Sure.
ROMAN DANYLIW: EKR.
ERIC RESCORLA: It's been quite a while, but I wanted to follow up on Martin Thomson's comments about AI here. So on AI-generated input, I agree, our defenses are proving inadequate. I know there's been a lot of discussion about AI-generated drafts and disclosure, but I'm actually much more concerned about the AI-generated mailing list traffic. The harm here is there's a huge amount of traffic that is superficially responsive but actually nonresponsive to wade through. If I send 10 lines and I get back 50 because someone hit that into Claude, that's not an ideal situation.
I know that Martin's observed that authorship means you take responsibility for your text, and that's not consistent with people behaving that way. I think we need some new norms here, and we're going to need the chairs to take a more active role when it looks like people are spitting out, you know, LLM output.
I'm not saying you shouldn't be using LLM, but I'm saying if you're the third person to have a message that says, here's the scene, maybe you should be asking whether you're generating your own text or letting Claude think for.
Secondarily, on the topic of AI BoFs, I share Martin's concern about the number of quality of the AI BoFs. There's a huge amount of interest in the topic and for good reason. But it's also a topic where we have a limited amount of historical expertise, and it leaves us in a weak position to evaluate which work is real and which work doesn't generate a process, doesn't really reflect wide-spread support but something else. This has been an issue both in terms of BoFs and the Dispatch. We had a Dispatch interim where basically everything just got punted because it was all stuff that was not going anywhere. And it's not a great situation. We burn a lot of meeting time to tell people no.
I think what is required here is some strong leadership and the leadership to feel comfortable saying, you know, I understand you want to do this, but it's not ready. And you're not going to succeed, and you have to come back with something better.
I know it's not a pleasant situation as an AD, but it's something the ADs and chairs may need to do. And I think, at least in my perspective, they would have community support for decisions.
I would encourage the chairs and the ADs to feel empowered to exercise their technical judgment, which is why they were appointed to the position, rather than just turning the crank.
ROMAN DANYLIW: Understood. Thanks for the feedback. On that first point, definitely, the interaction on the mailing list is part of what the IESG has definitely heard. We've heard it from the working group chairs. We've heard it from different director reviewers or how the back and forth goes. And in addition to identifying mitigations for that and with those associated norms, that would have to feedback to some training or summarization components so broadly working group chairs understood they have the resource they need when they encounter those situations.
On the idea of ADs taking more responsibility and ownership for how the BoFs are selected and doing their due diligence, I think it's a subjective call. I mean, your feedback is taken. Unless --
ÉRIC VYNCKE: And as the responsible AD for one of the AI BoFs, I always prefer to err on allowing the BoF, even if I'm not fully convinced to run it, rather than closing a BoF that could have been run correctly.
Now, on the specific BoF, of course, nothing is set in stone. The chatter that's been discussed will not be the chatter that would be going. As you know, the process is to get the community involved there. So, it's really up to people in this room and outside to work with the chair to make a good charter and possibly create the working group.
ERIC RESCORLA: For the record, I think that BoF was prematurely held. I think the BoF was not well organized and prematurely held. Of course, it's your decision, but I'm just saying that's one of the things I'm talking about.
ROMAN DANYLIW: Okay. Thanks.
ÉRIC VYNCKE: Fair enough.
ROMAN DANYLIW: Erik.
ERIK NYGREN: As powerful as the tool as it is, are there areas where we need to react more quickly from the IETF in a defensive measure? In particular, one area is IANA codepoints where their codepoint registries where the past things, like expert review, might have -- or First Come, First Served -- were fine when humans were writing drafts. But now as we start having AI-generated drafts, which codepoint usage could go up, go up orders of magnitude, is that something where the IESG is looking at giving out guidance to expert reviewers in the short-term so that we can get ahead of this?
ROMAN DANYLIW: To my knowledge, IANA is tracking in broad strokes what's going through the pipeline to include they regularly discuss to us some of the approaches, even for First Come, First Served, which could have some of the problems even before the age of AI.
In no way am I suggesting that tomorrow this may not appear. From the information we have now, this is -- at least right now not a phenomenology, not that we shouldn't consider it and shouldn't have some telemetry or oversight that would give us insight if and when that occurs. And we would hopefully have a mitigation lineup when that's happening. So, yes, I think that goes on to the larger list of considerations when we come up with the norms in this future forum. Good point. Thanks. Bron.
BRON GONDWANA: I'm over here. Is this live?
ROMAN DANYLIW: It is.
BRON GONDWANA: I have a couple things around norms as well. One of which is possibly a cultural change a little bit at the IETF that I think we need to make, which is identifying a human being. That in the past, you had to have a human being in order to get text into things. It's much easier for non-human beings to send things now. And requiring that there be an actual human being accountable for everything sent to the IETF, I think, is going to become really important, and we should get ahead of that.
And then following on from that is requiring that accountable human being to be in the loop. And in my newfound role on the Moderation team, I think it's going to be very important for us to identify the human has taken themselves out of the loop and blocked them from interacting with the IETF. And there might be rate limiting requirements we need to build through our processes to catch that before it goes crazy and cut those loops. Because otherwise, we're just one accident away from everything being wiped pretty fast.
So yeah, those two things. Identifying a specific human being who takes accountability and having the norms be that human being remains in the loop and doesn't automate anything that they don't check.
ROMAN DANYLIW: Understood. It'll be critical consideration in the future.
TOMMY JENSEN: So, it's come up enough times that I'll just quickly point out that there is an action item that a few of us are working on to write such guidance. You're not hearing specifics from us because it doesn't exist yet. So we want to speak with one voice.
So to everyone that's underlying in the chat, I'm sure I can't see, that is something we are exploring because we share that. To the point of real humans, note that we can use what we already require, which is that someone posts with a real name for definitions of real that we then need to be able to hold accountable. And that is something that we are considering.
BRON GONDWANA: I just wanted to insert, I saw something that looked like it could be not a real human being fairly recently that had a name and domain that someone had created for it, but it was a sock puppet. Sock puppets have always been a problem. We have a sock-making machine now.
(Laughter)
ROMAN DANYLIW: Jim.
JAMES GALVIN: Hello. Speaking of AI, you've all heard of AI, right? I have a question that I would like to leave you with as you think about how you're going to better manage work that's going on here in use of facilities and services.
So, three weeks ago -- it really was three weeks ago -- for the first time, I actually had a thought, a need that I would want to actually schedule a side meeting. I thought it was great years ago when we first added side meetings and we added a room. And then over time, that became two rooms. And then, we added this nice scheduling tool. Used to be a sign-up sheet on the front of the board and all that. I think all that's great.
But three weeks ago, I was thinking I wanted a meeting for my working group. We had a particular technical topic that I knew we wanted to take outside the meeting.
So I went to the page. And I'm thinking, oh, let me check this out. How do you schedule meetings? What does this look like? And I saw a few meetings scheduled there.
So then I started to look and think, okay, so let me pick slots that are going to work for me so I can offer to this group of people.
And then I went back to the page, and half the meeting rooms were gone. Half the meeting slots were gone. And then I'm just sitting there watching the page, and the slots are just filling up. I mean, this was three weeks ago. The meeting hasn't even started yet, and the slots were gone. And you look, and most of these meeting slots are related to AI things. Okay.
So my question to you, really, is I thought that -- we talk about the amount of AI stuff that's here. And it's interesting. We can certainly debate whether or not the IESG made good choices in the BoFs that were chosen. But more importantly, people who were not selected for a BoF, well, they've got it out here. What we have is a second IETF that's forming. And they're using the side meeting rooms to do it. And they're just having all the discussions they want to have that they can't otherwise get announced and available in a formal way.
So my question for consideration here is: Are side meetings -- what is the purpose of those rooms? I guess I had my own personal expectation that the purpose of those rooms was for actual IETF work that's going on that needs an opportunity to discuss something and get together.
I mean, maybe that's a bridge too far. I don't know. But I'm asking for some consideration. As you think about managing the work and use of facilities, should we think about whether or not those meeting rooms should have some priority? Should they be for actual working groups, actual work that's going on? Or I mean, they have basically been taken over. They're just not available for IETF work. That's the observation that I make.
And I ask you to think about that as you think about managing work. Is there something we can do there in terms of managing the priority of where those side meeting rooms go to? I had to find a different way to get the meeting happened that I needed. Could not use a side meeting. I mean, I managed it, but there you go. Thanks.
ROMAN DANYLIW: The IESG is intimately aware of the continued evolution of side meetings. In recent years, the idea that it is just extra space for working groups to have more organic, non-working group, you know, low-level conversations, it has not been like that for some time.
We have been watching and trying to find the right balance between this area where all we're trying to do is provide a room, which proponents say we need comms, and we need this and it became more formal and formal.
And we are trying to find the right balance that this is officially not working grouping space. This is not research group space.
We have actually analyzed the larger bucket of what's happening, and there certainly is what is the more historical understanding. But we also have proto-research group doing things. And we also have a lot of what we would summarize as pre-BoF storming and norming of ideas, and sometimes it does manifest into a BoF. I think several of the BoFs we convened have met a number of times as well.
And then we also have mini conferences that people are using the space for, in there to recurringly talk about things that will not, ultimately, make it way into the standards or the research process as far as we can tell.
So that's really to convey, we are not blind to the dynamics that are happening.
What is different actually about this meeting is that to my knowledge, this is the first time that we actually ran out of side meeting slots. In previous meetings, you may not have gotten your preferred slot, but there was a slot always available. And I think it's new information for us that we have reached the point where we have run out of slots.
The tension for the IESG is the more it puts its finger into this unofficial thing that we are not managing, the more ultimate responsibility we have. And if the signal from the community is they would like the IESG to interdict the kinds of slots that are taken and make value choices about what's in and what's out and provide that facilitation, we can certainly explore in that direction, but this is the tension we're managing to give the community the ability, this freedom to operate and storm and norm and then bring it into the process.
But if we ultimately hear from the community that there's interest, they want us to make choices, this one's in, this one's out, this one already met, nope, you don't get another one, or even talk about what's allotted, we could get there. Right now, the rule we have is you can't run a working group session in there. So working group chairs can't call a third working group slot or call consensus. That's roughly the line in addition to the Note Well applying.
We welcome more input on the community side where that balance is. That's roughly where we are right now. So we are not unaware.
JAMES GALVIN: So maybe my question then is to the community and to all of us here in this room. What do we want from side meeting rooms? I take your point that you're not managing them, and you haven't been to date. And all of that is fine. I get it.
If you're looking for input -- so one community member asking for consideration of how to better manage how these timely slots are used up. And the question is really to us, you know, really to all of us. If you want complete open flexibility, fine. If that's the consensus, that's what we'll do. But other folks should, please, take the opportunity to tell your ADs and the IESG, in general, if we would like to manage that better so that they're better available.
ROMAN DANYLIW: Very much. We welcome that. And the last one to your point, I didn't comment on the fact that there's a second IETF inside the side meetings, we are also aware of that and the tension people have, their ability to manage are they going to the working group or they going to try track potentially this new work that's happening and the timeliness of scheduling relative to them. We are aware and we are looking for feedback.
All right, I see Andrew next in the queue.
ANDREW CAMPLING: Yep. Hi. Coincidentally, I came partly to make the same point that Jim just made about the AI task force work always seems to meet at the same time as the IETF.
As a suggestion, I think it would be helpful to at least have some very light touch curation to ensure there's space for other topics without getting too heavy handed in it. So that will be at least maybe a starting point.
Conversely on the AI BoFs, I'd push back against the comments made earlier and say I think we shouldn't make perfect enemy of good, as the saying goes. Yes, you could have been crafting those BoF briefs, whatever, for the next couple meetings by which time it will be irrelevant to have held them. You know, because we're just not moving at the right speed. So I think it was absolutely the right call to do something now rather than do something better later that was not relevant, then who cares what we think.
And final point really briefly, said earlier, but I think we all owe the TLS co-chairs and the AD a vote of thanks for still bothering to turn up given the appalling behavior that they've been at the receiving end of. So thank you.
(Applause)
ROMAN DANYLIW: All right. Thank you for that feedback, Andrew. Tara.
TARA TARAKIYEE: So, first of all, plus one to the solving the side meetings problem. I really missed not having the Public Interest Technology group meeting this time.
But the main comment I wanted to say up here was just to, I mean, thank you all for taking the issue of diversity seriously. I think there's something we talk about nearly every IETF, at least that I've been. But I think at this point we have to stop talking about it as if it's a pipeline issue. Just simply telling people to apply will never solve this. It's a structural issue. It's not the compositions of the different groups here are not accidental by-product of the way we work. They're a result of it.
To paraphrase something said this morning in the GENDISPATCH about the draft on NomCom gender composition, like you can't at the same time want change and then be afraid of it.
I think there's lots of smart people here. We can figure out how to change the structures so that it results in a much more equal representation because, I mean, as we say, like the Internet is for end users. It should represent -- the people making the decisions about this Internet should also reflect all end users. And yeah, I mean and forgive me if I'm being a bit incredulous seeing how quickly the community has moved around AI whereas this topic has been discussed meeting after meeting. Lots of interest in solving the problem but not lots of movements towards doing the running code of solving it. So thank you.
ROMAN DANYLIW: Thank you for your intervention. Usama.
MUHAMMAD USAMA SARDAR: First I'd like to have a few topics which is starting with the BoF. Like the AI BoFs, I think many people have commented that the BoFs were not good. I had like a very basic question on how is that AI different from any other workload that we have worked in the IETF, like all the protocols, all the work that has been done? How is that AI that we are all talking about, not only -- I don't mean only academia or industry itself, like all people involved. Like, what exactly is it that makes AI different from any other workload, any other protocol that we have used?
That simple, basic question hasn't got any answers or any, let's say from my perspective, any response or any answer on what exactly is it that we are selling with the name of AI specifically?
But there still remains a lot of interest, and I have seen in the side meetings and also there was a point about side meetings. A lot of side meetings are just happening because there is no AI working group, so to say with the label of AI. People want to do something, but they're not being given a working group, so to say or a research group. So things are happening in the side meetings that people are somehow trying to do something, trying to figure out what that is. But I think the very basic questions still remain unanswered in this form, as I have given an example.
The second point I want to make is that also was mentioned about AI, people generating the drafts using AI. It's very easy to generate a draft now, but it's also putting a lot of workload on the reviewers who are reviewing the drafts, in a sense that like with one command, you can generate a draft, but now somebody has to put in a brain in order to understand what that is, what these terms are, and what the draft actually is trying to achieve and so on. So that workload shifts from drafting or authoring or editing, whatever, to reviewing that workload balance needs to be somehow managed. At least in the academia, we have a thing that if you use AI, you properly acknowledge it somehow in acknowledgments or wherever. I think we need these best practices also in the IETF, at least that people who are reviewing the drafts, they know that somehow some part of it or a substantial part of the draft was written using AI. And the authors, even if they use that, other than the acknowledgment, they really have to take responsibility that AI was used. And any mistake, any vulnerability, any kind of problem, they are responsible for it and not the AI itself.
The third point I want to make is also somehow related to the side meetings. It was said in an official meeting of the working group that consensus has been developed in a side meeting. That's completely wrong, as far as I understand. Side meetings are not supposed, as currently the instructions go and the best I can understand currently, side meetings are not supposed to have the working group business being held in there. And please correct me if I'm wrong, all of you. And it was exactly stated like that in a working group official meeting that consensus has been developed in a side meeting that was held two days ago. That's something that blows my mind. And that's something we, as a community, need to understand that side meetings should not be used for the working group business, and consensus is to be developed within the working group business meetings and not in these side meetings.
The fourth point I want to highlight also is that there are no processes in the IETF, the best I can understand is, about the vulnerabilities for the Internet-Drafts and also for the RFCs. If some vulnerability is found or by some process, some implementation is vulnerable to the implementation of some Internet-Drafts, there has to be some process somehow that we can indicate somewhere that this draft is vulnerable to xyz, or xyz, blah, blah, blah, that I think should be a part of the draft itself and the security considerations or somehow it has to be incorporated.
I
was told that this is independent from what we do in the standards body, but I don't believe so because an implementer has to really get a feeling of what is the security risk that he will be vulnerable to by using -- by implementing an Internet-Draft, which he's reading or is going to implement.
So these are my four suggestions that I would like to request. Thanks.
ROMAN DANYLIW: Thank you for the feedback. To answer the one specific question, yes, consensus is not called in a side meeting. It is not part of a standard process as you mentioned the other speakers.
I observe with the queue is continuing to grow. We do want to get feedback for the open mic. We are soon to be over time, so I would just ask for all those coming up to the mic to be brief and pithy in their point. And we will equally try in our response to be responsive but brief.
Next up, I see Peter.
PETER LIU: So, about the time consumption of the AI discussions, I agree that we could have better total time control and quality control on discussions. But actually, the side meetings, I see the gentleman discussing the purpose of side meetings. I think side meetings BoFs are instruments for IETF to understand and discuss and potentially incubate new work.
I think right now, side meeting does not represent IETF standings or does not go to official proceedings. The working group has their own resources to publish their own consensus and everything. So when you look at the total time consumption about AI discussion, there are four BoFs or three BoFs out of the whole IETF week. There was like 130, 150 total sessions. So that only takes up like 5% of the time. While as a new trillion-dollar industry, I don't think that's really, really blocking the IETF pipeline or anything.
Many working groups in the history are somehow experimental, although they try their best but failed, but at this time, some side meetings or some BoFs that are doing at least I see the chairs and proponents are trying their best to present a problem statement and a gap and everything, I think those are these things that -- my message is simple is that we should at least allow discussions and opportunities at side meetings and BoFs. And of course, with a good quality. Thanks.
ROMAN DANYLIW: Thanks for the feedback. And a comment on your stats on the number of session requests, you roughly got the size exactly right.
Okay. Nathan, Nathanael.
NATHANAEL RITZ: Just a comment. I saw Jay Daley mentioned that there's an email I can send comments to as well. Just for the side meetings, the biggest challenge that I found that was trying to follow through on the link. It's going to an unrecorded Zoom meeting or something like that. The slides that are shared aren't easily accessible. And I think that's -- I think these side meetings could have more value if there was more organization when it's being presented as, you know, here's a link on the IETF website. The side meetings, here's a nice interface.
In that process, if there was a way to have resources made available the same way as the working group meetings, that would add incredible value. I know it adds additional labor and work, so there's a balance there. But that's the biggest challenge that I've had working or trying to interact with the side meetings.
ROMAN DANYLIW: Thank you for that feedback. I'm going to catch that one instead of Jay. There is a WebEx that's consistently provided for each of the side meetings. The distinction is it's not the Meetecho, the overall standards process vehicle. That is an explicit design choice to separate the things that are inside the process versus the things that are outside the process. But there is WebEx available for each of those meetings.
SEAN TURNER: Hi. Recovering finance person from the LLC Board. I would like to provide a solution to the side meeting issue. If you would like to conflict with an IETF working group meeting, it should cost you some money. At the end of the day, we can decide what it is. But it's good to go. If it's not a conflict, it's free. And we'll figure it out. I think we’re in this problem where it's a simple supply and demand problem, and it's free, so they're going to fill up. If you want to conflict with the actual work of the IETF, write a check.
(Applause)
ROMAN DANYLIW: Thank you, Sean, for your feedback. Ben.
BEN ROBERTS: Yeah, hi. I'm actually completely new to IETF.
ROMAN DANYLIW: Welcome.
BEN ROBERTS: Thank you. But this aspect I'm hearing a lot about emails and AI flooding email lists. I'm on the Board of AFRINIC, the Internet registry for Africa. And we've been having exactly the same problem on the policy development group, and it's gotten really bad and really sophisticated. And we are also trying to work out what to do. So we share experiences as we come up with solutions, but it's become a real serious problem with people, you know, really annoying everybody else who is actually a community member, while it's not just about flooding messages and repeating messages, it's actually making things unpleasant.
ROMAN DANYLIW: We look forward to the ability to share our practices. I can't say best practices, but as they evolve, the same thing, I think we look forward to doing that with the ITUT. Thanks again for coming.
PHILLIP HALLAM-BAKER: Yeah, the AI craze is recent, but AI is not. As you go, the use case that keeps being thrown at me, well, I have one AI here, and when it talks to that AI there, they want to drop into AI speak rather than have to express everything in natural language, which is great, only, that's exactly what Tim Berners-Lee was going on about in this thing called the semantic web for 30 years. So I think that maybe what we should do is make them read his book first.
(Laughter)
ROMAN DANYLIW: Thank you for that feedback. I observe that has drained the queue, and we have the LLC behind us. I will ask one more time, are there any other questions for the IESG? Someone going to the mic? No. Someone is leaving. Okay. Have a good night.
(Laughter)
No worries. Have a good night. All right. Very good. So IESG, thank you. And coming up to answer your questions is the IETF Administration LLC.
(Applause)
MIRJAM KÜHNE: Hello again. This is the LLC. Board Chair, Roman, board member, Jay, Executive Director, that's all we've got this time. So any questions, comments?
ROMAN DANYLIW: Rohan, we punted you. Could you come back up to the mic and give us your question?
ROHAN MAHY: Hi. I came upstairs to change. We have an election in the United States ostensibly on the 3rd of November. That's 11 days before IETF 127 starts. What mitigation plans, if any, do you have? What do you anticipate being able to do for IETF attendees in, you know, a variety of bad to terrible situations?
JAY DALEY: Hi, Rohan. So we have nothing planned for any specific scenarios. That's not anything we would do anywhere near at this stage. You know, no need to be in the situation rather than guessing there. We do, however, have a BCP produced some time ago which gives us guidance on how to manage emergency situations. And the timing of those emergency situations out from the actual meeting. So, if anything were to happen that actually was something happening rather than being feared, then we would follow that BCP.
MIRJAM KÜHNE: Thank you. Next in the queue. There's Jonathan then.
JONATHAN HOYLAND: Hi. So I was wondering, just as a sort of follow-up to the previous question, what are the plans for keeping people who practice Judaism safe in Malaysia given it is a criminal offense there? And as separate from being Israeli, which seems to come up a lot, given that I'm not Israeli.
JAY DALEY: We don't -- as I think you're probably aware, we don't make plans of that, for the safety of individuals in that way. We haven't done that in any country in any IETF meeting. It is an individual choice about coming to meetings. If there is – someone can come to a meeting. We understand there are oftentimes people who cannot come to meetings.
The bigger picture answer to your questions, going to the heart of it here, is that we live in a difficult world. We have a very, very broad set of IETF participants, and finding venues that everybody can be comfortable in every time is impossible. We have to, therefore, to be as fair as we can in distributing the pain
Now, one of the biggest issues we have and repeatedly have had is about Asian participants unable to get Visas. I'm unable to travel to the U.S. or North America or to Europe. We also have a requirement to meet in Asia. So, I understand that this is very difficult for a number of people to go to, as I understand Singapore was very difficult for a number of people to go to. That is a reflection of the world that we live in, where there are different norms.
Now, I don't mean to say that in a way as to devalue and suggest that there are, you know, some sort of moral equivalence here. That's not my point. But we do have a situation where we simply have to distribute the pain because there is no other solution that we can achieve.
JONATHAN HOYLAND: Was there a reason Bangkok was not selected, given that it's very close by and much more friendly?
JAY DALEY: We -- there are -- we have requirements and requests for multiple different countries because of the way that Visas work and the way people can travel to things. We are going to look at Bangkok again, definitely. All right. But it is -- we cannot concentrate simply in single countries, that every country we choose has problems. That's why we have to distribute it.
MIRJAM KUHNE: Thanks for taking that, Jay. I think there's nobody else in the queue. I think the IESG drained all the energy.
(Laughter)
Well, thank you very much then. Wait, wait. What? Oh. Martin. Sorry.
MARTIN DUKE: I would like to endorse Sean's idea about side meetings. Side meeting rooms are actually one of the few things we do that have a high marginal cost to the IETF, and if there's a lot of demand, it would be great to get more. But that costs money. I think just having some sort of -- I don't want to say nominal fee, but not large in the scale relative to the attendance fee here might help solve the problem. Just by getting more inventory. And also constraining demand. Thanks.
MIRJAM KUHNE: Yeah, thanks for the suggestion. That was it, right? Am I missing anyone? No. Well, then, I wish you all a good evening. See you tomorrow and next time.
(Applause)
Transcript Provided By:
Caption Pros
10105 E. Via Linda, Suite 103 PMB4
Scottsdale, Arizona 85258
www.captionpros.net
(With edits from Cindy Morgan)