shmoo WG Agenda

IETF 112 - Virtual
Friday, November 12, 2021
12:00-14:00 UTC
Session I

Chairs:
Mallory Knodel
Suresh Krishnan

Agenda

  1. Agenda Review, WG and Deliverable Info (Chairs)
    10 minutes

  2. Guidelines for the Organization of Fully Online Meetings
    Martin Duke & Mirja Kuehlewind
    15 minutes
    draft-kuehlewind-shmoo-online-meeting

  3. Running an IETF Hackathon Draft update
    15 minutes
    draft-ietf-shmoo-hackathon

  4. Recharter summary & discussions

Notes

draft-kuehlewind-shmoo-online-meeting

Alex question: Direction considering four slots instead of three?

Mirja: Four slots per year or per day?

Alex: There is a number 3 in this draft, instead of 3 to make it four.

Martin: RFC 79xx spcifies three regions in which to host meetings. Our intention is to keep a fair balance between those three regions. We could have any number per year and use this algorithm, but we're basing it off of this RFC.

Alex: 3 meetings per year, but cycling through 4 timezones (not time slots).

Martin: We didn't consider that beyond IETF consensus document that we should attempt to balance across the three regions listed.

Alex: There are regions and there are timezones. Some of the regions have more timezones than others. Asia has more timezones than the others.

Mirja: We took the median of the timezones for each region. If we weight each region equally then there are only three values that go in the calculation.

Andrew: For scoring purposes you could alternate the Asia-PAC timezone rather than taking a median, alternating between India and Japan for example. The reality is you're going to end up with a variation of three start times roughly 8 hours apart with whatever algorithm you use. You could make the scoring more sophisticated but you might not end up with any better answer.

Martin: We played with it a little bit so if someone has a better algorithm then we're open. There's a theme here that maybe when we do North Am night and EU night slots, maybe for Asia we try to minimize the pain for India or Japan, that's hard to work through but we could probably do that.

Mirja: We can make the scoring scheme complicated but the result wouldn't improve. Really only have one slot that is really bad and so that the other two aren't so bad. We suggest for everything we considered that maybe one suffers the most but it's not so bad for the rest.

Martin: India experiences that there are two slots that are kinda bad instead of one that's horrible and two good slots. India is big and this is a hard problem but this is where we ended up.

Mirja: The goal is that you only have to suffer for two. Check your timezone and see if this is achieved and let's move on.

Nick: There wasn't as much advice in this draft about how to handle socializing. I'm not making it to Gathertown. Should we provide some recommendations for what to do?

Mirja: We mention that online meetings are more open but we're missing the social component. It encourages experimentation. I don't know what to add. Proposals welcome.

Martin: Open to proposing something to this draft. Might be limitations at home when the timezone is bad or you're at home. As opposed to when you have nothing to do at the hotel. Let's add a section to the draft but I don't know what to say.

Nick: It's going to be hard so let's force it a bit otherwise people won't do it.

Lars: 1/2 hour breaks are too short and too long. Long enough that I do other things but then I don't hang out on Gather. We might describe a 1 hour break.

Mirja: Draft says we shouldn't extend more than 6 hours for the day. We say to experiment and what we've done.

Lars: We can't eat into WG time so yes we have to extend.

Lars: Fully online meetings will go to hybrid and so discuss on the mailing list what guidance we want to give for these meetings. Should we use this scheduling for in-peron meetings? Flexibility for on-site but not terrible for virtual participants?

Pete: Gather is my home base and I use the doors to go into the sessions/meetecho and some experimentation around using the social space as the primary space as a potential place to increase poential for social interactions as they switch between rooms.

Mirja: Sounds like tooling.

Pete: Center around sessions or centered around social?

Mallory: Doesn't work for newcomers, day passers and invited speakers. You'd have to have the agenda as an alternative. And then everyone would use what works for them (the agenda).

Mirja: Let's not prescribe that in this document.

Mallory: Local meetings.

Mirja: The original idea behind many couches.

Suresh: Mallory to add some text.

Ted: Boston meetings didnt' have enough energy to happen consistently. We had off-meeting presentations that felt like work. Needs to be synergized with IETF not additional.

Ted: IETF meetings are just meetings on my schedule now. It feels like extra meetings. When I was there in person, taking a work meeting was a bit imposition on my schedule. It feels like the opposite. Gather isn't solving the social because I have to go somewhere to interact with people and it's optional. I have other stuff going on. It would be nice to have a way to have impromptu discussions forum that replicates what we did at in person meetings. Side meetings were happenstance. You have to spend a bunch of time in Gather. Lower-cost way to be in the same room with everyone rather than Gather. If Jabber were functional let's just have a jabber room.

Mijra: It's not useful to have an RFC for these low level things. High level things like you're raising need discussion and need their own drafts.

Suresh: Ted you can write a draft.

Andrew: Fundamental problem with social-remote is that we're remote. Completely replicating the whole of the experience remotely is probably impossible if we're all in different timezones. It's the wrong thing to try and replicate. It's not solvable remotely.

Andrew: When we go to hybrid meetings should we use the timezones in this paper? I think that's horrible. Surely we use the model where we use the timezone of the place where we are. We're asking people to be somewhere in person to have a dumbed down experience to suit the people who aren't there. Better thing to do is to upllift the remote experience to make it better rather than downgrade the inperson experience to make it worse. IGF if hybrid next month. Learnings from that would be nice.

Andrew: Big hotel chains has a deal with Zoom to faciliate hybrid meetings. We can go multiple hybrid meetings hotels will be sufficiently able to handle it.

Mirja: Concrete proposals please write it down and we can discuss based on text.

Andrew: Need a charter change.

Mirja: Charter is changing so go ahead.

Mirja: Additional benefits for remote people is a high level prinicple. Before the crisis we said we optimised for in person because that's what you pay for. We tried to improve remote for this crisis so should we continue those to the detriment of in person experience?

Andrew: They have a single mic queue at the IGF for example. You have to use your device even if you're in person.

Mirja: Everyone should sit with your laptop in the room insead of having a screen in the room-- this is an extreme example.

Dhruv: People active at the India hub were in the IETF meeting so it was difficult. If there's an India-specific travel restriction then possibly it will work. How do we organize hubs when interests are so diverse we needed 6-7 rooms. We do meetups, RFCs We Love. In hybrid, organize some event in Gather to make people come there. HotRFC was a good thing. Even if WG chairs could do something in Gather.

Mirja: Super to have hubs where there were actually meetings during the meetings. Write it down.

Mallory: Could put text in a section on Hubs.

Mirja: Happy to accept text but might need its own draft.

(in closing) Suresh: Adoption call for this draft will kick off next week on the list.

draft-ietf-shmoo-hackathon

Suresh: Goal is to get this draft out as soon as possible.

Recharter

Charles: Several people commented is it ready or do we need more?

Mallory: I think it's ready. But will be sent to the list for more discussion before it's finalised.

Mirja: Maybe we can discuss the changes here?

Suresh: Remove COVID stuff and make it more generic. Talk about hybrid. Meeting fee draft into BCP. Changes here: https://docs.google.com/document/d/…5juI6qPH_4FSKSCCUc/edit?usp=sharing. One month where a lot of people commented. Cool off period before we send it to Lars.

Mirja: Is how to incentivize social in remote in scope or not?

Mallory: If it's not explicitly out of scope then that leaves room for experimentation and ideas. Not sure it needs to be explicitly mentioned or it feels like an imperitive.

Nick: Eliot's draft suggests that the LLC should investigae the in person meeting cadence. This would be in scope according to the charter. I hope we could emphasize that.

Lars: It's in scope. The IESG an LLC had a workshop and the feeling is the pandemic and long string of online meetings it's not something we want to do right away and instead are trying to go back. Not wanting to change the cadence now. What would that look like? Would we replace one with a full online meeting? Would we add online meetings to the year? Adding meetigns would make the year too busy especially with interims. If replacing which one and why and spacing in person out further would be difficult to do. We decided to punt as we ramp back up from COVID.

Nick: Remote people aren't second-class and let's not roll that back.

Andrew: Why is it important for the people who go in person not withstanding improving the remote experience. Remote experience should be better than it was previously.

Andrew: Carbon footprint of in person it would be better to look at the footprint of its protocols than the few people attending meetings each year.

Suresh & Lars: Recharter will cool off for one week.

Jay: People strongly prefer in-person meetings and are experiencing adverse effects about online meetings.

Jay: It's not more expensive to run in-person meetings. They're cost neutral.It's more expensive for attendees and sponsors (who are fine with it.)

Jay: We've started doing carbon calculations and we have a process for this when we go back. We're looking at carbon offsetting and we haven't done anything about it because we're not going back, so we should be able to do that within the cost of each in person meeting.

Jay: A lot of work is going in to prevent the remote experience being second class. A new version of Meetecho is coming such as a single queue and those things. That comes down to the behaviour of the people on site. Are we going to do 6 hours or 8 hours and those are the things that will make a different for remote participants.

Phil: Yes I want to meet in person, but I don't want to do it more than once per year. When my kids were younger I couldn't do more than two per year. We assume the in person meeting structure works and it doesn't. You have to read the drafts. I do a least 12 working groups. Reading the drafts for 12 WGs is a huge undertaking. Most people can't do that. The IETF meeting structure doesn't relate to the way we do work as engineers.

Phil: On carbon offsets, the reason you make a symbolic gesture on carbon emissions is to remind us that it's important. We've never talked about the carbon costs of our protocols before. In terms of how we do our own work we do carbon offsets to remind ourselves that it's important. Kerosene consumption (jet fuel) over half goes into lighting. It's expensive, dangerous and poor lighting. If we gave everyone LEDs and solar it would have the same effect as converting every jet.

Christian Hopps: Our work has a multiplicative effect even if we burn some carbon. Reduction of carbon across the world because of our work is overal a net positive so let's not get in the numbers.

Christian: I find three times per year effective; we're not all the same.

Eliot: I don't feel as effective when I'm jetlagged. My kid said to me I was the problem with carbon. You can't say our online meetings are always beneficial. We have to find a balance.

Lars: Do you have a concrete change in mind that we can get consensus on?

Eliot: Collect a number of approaches that we could try, not standardize. We don't have to try them every year. We could plan a schedule for trying things. 1) semi-annual meetings for plenary and load up with interims that are remote. 1a) exend the meetings on either side by a couple of days. 2) Do semi-annual meetings but require blocks of interim meetings that occur in person on either end, eg March/November schedule would require interim blocks in July/December. 3) Go annaul and make them 2 weeks and do a lot of interims. The intent to ask for data from Jay is to help calculate what's beneficial. What's most important in the process that contractually the LLC sees us being interested in doing this so that they can manage the hotels.

Jan: Beside what's best for the environment, my employer Cisco announced that they will cut travel expenses 60% and ericsson is doing something the same. It won't be possible unless its continental travel. Everyone has to participae equally if they're remote.

Charles: We need to keep in mind that there's a spectrum of participation level. The real thing to focus on is the hybrid approach and looking to see how we can improve remote participation. As a result you will have more remote attendees. We make that decision every time how much time to devote. If remote experience is good they will use it more because they can't or don't want to travel. We need to make sure we don't mess up remote participation such that everyone feels they have to travel. Or we make in person so bad that no one shows up. Give that hybrid balance some time and then we can meet in person less often, maybe.

Andrew: Interim based schedule feels divisive for the community. Makes it harder for people to participate frequently because the choice of the meeting time is based on majority. Should remote interims use the guidelines from this group's draft?

Martin: The real difference between being remote only and in-person is side conversations. We can't cut off the benefits of in person to suit remote.

Martin: If companies are restricting to continental travel then that's why we rotate.

Martin: IETF 107 was emergency situation for QUIC and we had a month of meetings at 17:00 UTC and that wasn't a good model. Wouldn't be great to drive a lot of WGs into interim mode.

Suresh: IoT cluster is like a mini-IETF and they're integrated technology wise. The real answer is that it really depends with interims. We have to make sure that different needs are accommodated by whatever we do.

Mirja: We touch the interims in the draft. The dynamics are different than in person. They achieve different goals. It's much better to have small but deep discussion that is interactive. But meeting week you have more particpants that aren't as close to the material but have useful input. Use both meeting forms appropriately and to consider if you need to meet at the in person meeting or if you can do interims. The draft discusses this that it depends on chairs' preference. Happy to take text proposals for the draft to discuss this more.

Suresh: Closing the meeting!