 So this is all my writing here. So it will be intensely boring if I just talk. So let's leave the agenda aside for the moment or leave this document aside for the moment and start with questions. If people who are here who weren't on the team this year or Gregor, you still have some questions about bursaries, then so let's open up to a question. Is anybody in the internet? Luca, hi. You need a microphone because we're sharing. Could you explain for those of us who are not part of the bursaries team the impact that cancellations have on you or the budget? So on me, it's just more work. It's basically so I send tools and it just means there's more batches. And there's typically some sanity checking back and forth with the rest of the team, which means there's extra males on the bursaries list in the sense that we start with the highest ranked candidates in some sense and work our way down. And so the question is, okay, guys, we have some more money. Does it seem reasonable that we fund these people who were ranked like this? I don't know, is that an answer to your question? It's not a terrible amount of work, but it seems to sort of be a task for 45 minutes or an hour a day during this period where we're dealing with it. Do you have a situation where you have cancellations that are fairly late and the next people on your list to whom you would have offered a bursary had you had the funding originally are unable to accept it's too late in the schedule, that kind of thing? That happens more rarely than I expected, to be honest. But I think that's one of the things that I mentioned here and we might as well talk about it now, which is the way we do things, the way that I've been doing things for a couple of years is to take the budget as a strict limit and never have more offers than that outstanding. Given the reality, we have a very strong expectation of cancellations and we can probably look at the data from previous years and every conference is different, but still we can, I think it would be reasonable to go to the DPL and say, this is our budget, we're going to overallocate by 10% or just to pick a number out of the air, but we expect to be within this lower number to give that kind of slightly more complex budget. I know that previous DPLs when we discussed that idea with them, so I think this when we talked to Luca, he said, I have no problem with funding it, but come back to me when you need the money. And I think, this was a little bit different because we were talking about working our way further down the thing, but it's closely related. I think that if we can offer more people earlier, it will be more efficient for everyone. I mean, less people will have to deal with last-minute panics. Yes, we'll have less rounds. So I don't know, maybe if you have any idea as DPL in the current hot seat, what do you think about this? Hey, hey, this year I guess we increased the budget for groceries by maybe 30%. Right, which was more like, that's actually another topic. So one topic is that we had this pre-process discussion with people and I said, okay, with this amount of money, we're only gonna be able to fund travel for about 45 people, that's my estimate. Is that okay or should we go to the DPL now and ask for more money? And well, I mean, everybody's in their own silo, right? And very stressed, so I didn't get any feedback that said, in fact, the first feedback I got that the number was too small was from you after we'd gone through this. So, I mean, it's not your responsibility to give me that feedback earlier, but somehow we could have also made things more efficient by asking for that extension earlier and looking at things and saying. Yeah, before making public the first results. Yes. Yeah. Very supporting of the idea of the, for example, the 20% extension over-committing, yeah. Because we can afford it. Right, even in the worst case where nobody cancels. Yeah. Right. Yeah. And for me, there is another question which is about the debt budget. There are things that should be constrained a lot because they cost a lot, like the venue, possibly. The accommodation, the conf dinner, whatever. And I think that the budget for bursaries should maybe be distinct from the debcoff budget because it's mainly with Debian money, maybe not. I guess, I don't know. I mean, it's, in practice it hasn't been a problem because Cathay wants to ask some questions. No, no, it's okay. You want to ask something, too. But you can. So, at least this year, there were no, there was no difficulties. I mean, I guess the budget is, you know, the lot of. What I wanted to say is that we can increase, no, that's actually false because if you have too many people for the venue, you're screwed. Right, so one thing I tried to do this year is track better the sponsored accommodation because it does matter, right? It's 25% of our budget this year is sponsored accommodation. And it's not an expensive venue for accommodation. I take it off. All right. That's why we have discussions, right? Sure. But having the budget together, Debcoff and Borsari, has given some concern to local team because they are negative numbers. So, I think they will be more relaxed if we split and they don't need to balance the, also the Borsaris. Okay. Reminds me of my initial remark. People tend to focus a lot on the deficit of Debcoff. And this year, a big part of it is because of my decision to raise the Borsari's budget. Right. They might think that they were doing bad fundraising, which is not necessarily the case. Right. I mean, from my point of view, it's a question of how you present the accounts. Yeah. And so, whether this is a big change or just a way in which the budget is written up in terms of accounts. You were talking about separating the venue and the Borsari's budgets out. But doesn't the two interact so closely? Because if you're in a far away place where there's lots of travel required to get Debcoff, we're going to get a lot more people asking for Borsari's because it's beyond their normal budget. And at the same time, those far away places may well be a cheaper venue to have than if you're in a more populated area where the travel is cheaper to get to or the accommodation is easier. So the two really can't be taken separately. So part of it is the only way we do fundraising right now is through, or active fundraising, is through the Debian Conference sponsorship program. And that's part of why Borsari's has been included. I mean, there's no reason, like Borsari's is something we know we're going to have every year. We know we want to send as many Debian developers as possible. And that could, I think, if we were doing sort of global fundraising for Debian developer travelers, I think that would be fine and should also cover the mini-confs as well. Not just the one main Debcoff. But for now, I think, yes, it makes sense to keep it in the same budget because we are fundraising it for all as one block. It's okay that people are a little bit stressed about fundraising. Yes. I just wanted to, I walked in late. I'm sorry about that. So please tell me that it's been discussed. But when I walked in, you were talking about putting more diligence into the sponsored accommodation. Because in DC-15, that was like difficult to do. And now for DC-16, I think you improved the process and you had a closer watch. My question is how much do you think would it facilitate the entire process if we didn't have the concept of sponsored food and accommodation? But our attendees who are sponsored would simply receive an invoice. This is your invoice for attendance of Debcoff. You chose to be here five days at $30 each, so $150 is payable. And then they take this invoice and give it to you. So this means running more money through the trusted organizations or not? No, I think how you do the treasury, that's something separate. But it's more about creating a, lacking the word, an interface between... Awareness? Well, that too, that's a huge part, but it's also an interface. For instance, that would be compatible at some point in time with taking accommodation out of Debcoff, Oregon, entirely, if we ever wanted to do that, so that you put your promo code into that hotel's website over there and book your rooms, and then you apply for bursaries to give you the money back or somehow get a charge to a bursaries account so you don't have to shift the money all over the place. I'm simply wondering like, would it make more sense to simplify the interface between the applicants and you in invoice form? What I would say is no, and the reason is no is because I just about lost my temper today with people showing up at three minutes to two in my 12 to two document collection block, and that's with only travel sponsorship people to deal with. So the fact that food and accommodation sponsorship, which will probably continue to be a much larger number of attendees doesn't involve any time from bursaries slash treasury people at Debcoff is probably a strong benefit that I would not wanna give up. I think you wouldn't need to have this be a reimbursement process for the accommodation because all you need to do in advance of the conference is say this person applied for $150 for accommodation and food and then by checking that off you're essentially diverting who's paying the funds. It's no longer a receivable from. It could work, but the processes for dealing with the venue every year are very different and usually weird. Which is why I think the attendees should be dealing with the venue and we should have a standardized process. I think it could be difficult because then we are buying our attendees and for tax reason and trusted organization it could be very difficult. If we're in both it's okay, but if we pay and then they should come and I think it's more difficult for taxes reasons, but it's an idea. Maybe I could just, it seems like an implementation detail which is fine, we can talk about those but maybe since we're all here and this is probably something that the person doing it in a given year can decide. Probably the community as a whole doesn't care by their way, so maybe we could stick with it. I don't know, if you don't mind tabling it for now. And then I killed the discussion. All right, so other things that people, other questions that people have. Fotis, is your name, sorry? Yes. Fotini, I'm sorry, I get your first and second name mixed up. You're just a spectator or did you want to ask some questions? Okay. At least I learned something. Now I can hopefully remember your last name since I've been publicly embarrassed about it. Okay, other questions. Luca again, these Canadians, I do not know if they're worth the trouble. When you said that you track the accommodation, what is involved in that and why is the accommodation tracked? So previously we have not done any, so for example in Heidelberg it was very different because the accommodation we bought in bulk. Is that right, Martin? More or less we booked a certain number of rooms. We booked the venue. We booked the whole venue, right? And so at that point the incremental cost to Debian of giving food and accommodation sponsorship to people until the venue filled was zero. So there was no real accounting to do. This year we pay per person per night. And for example, when Manny approached me and said, hey, paraphrasing, should we sponsor more people for travel? I said, sure, but we have to also take into account this will cost extra accommodation and made some estimate about if we sponsor these 10 more people then it's gonna cost this much more in accommodation. Again, having come late to this discussion, please tell me to read the documents or whatever it is. I have a couple of questions, a couple of points. One of the things that I see there is the always cancellations overcommit by X percent. So we agreed that that was a good idea. Okay, and I'll shut up. I do wanna say that it worked really well this year. I think in the sense that when there was a cancellation you guys, you were really quick at pulling in the next people. And if that process had been stretched out to a couple of months before, I think it would have been highly successful. So I think that was really, really good. So I could just comment on that. The reason that that was possible is that we did one static ranking initially and used that ranking for, so we didn't have to have discussions of people's ranking in order to reallocate them. Right. Did you discuss about what went well in the process and what didn't? No, we've mostly been complaining, so. So I had the impression maybe just because of who was on the diversity team. So let me not forget anybody, Gregor. We had Gregor, Hogar, Karaj, Sledge, Laura, Tiago, Martin. Sorry, Martin, I keep doing this to you. Martin really was helping. I think that's every, did I forget anybody? All right, so if anybody was forgotten, feel free to curse me out on IRC and then Nicola will be happy to read your curses out. See, he's nodding. French curses preferred. Okay, so I think that group of people for whatever reason was good at thinking about different ways that people contribute to Debian or people who had contributed some but showed some potential or were bringing some diversity to the conference. Not all of those people that we offered managed to make it. There were several people that were funded that canceled several people from Africa that we funded that canceled for personal reasons. It's not perfect but so that I thought worked well. I think part of the reason why it worked so well was because you made the decision to not let everyone rank everyone but you had two people, two from the bursaries team on each record so that there would be like some sort of weighted opinion and that as a result meant that for the entire process of ranking like however many people there were, I don't remember, it took me less than an hour. Because I knew most of the, maybe I was lucky, right? But I knew most of the people on the list and I didn't have to do too much research but it was really not that much trouble and so next year we can definitely tell this as a success story and hopefully get even more people because there's gonna be even less work and to get the static list which was really, really helpful. So one comment that we were discussing previously, I think during the cheese and wine party was at some point I suggested or some of us agreed that it would be good if people applying for travel sponsorship were not on that year's bursaries committee and I think everybody agrees this is a good principle. The conflict of interest is pretty obvious even if you recuse yourself from specific cases, the fact that you're involved with discussions of other people that the conflict is obvious. It does make it a bit harder to find people and this year was particularly challenging because so many DDs reasonably felt they needed help getting here which Andy already mentioned. So it's a reason, I think that more than the workload made it hard to and maybe I was a little pessimistic about the workload. I don't know Gregor, do you mind saying how much time it took you to do or what was your experience? So my experiences are quite similar to what Martin said since basically you were running the show in a perfect way. We just had to open up this spreadsheet and go over our, I don't know, a dozen or whatever person. Yeah, it was more like three dozen, but. Three dozen was okay. I didn't know it. Yeah, so I guess the first batch was, you said an hour, yeah, maybe an hour and a half or something like that. And then those two or three additional persons. So it was really not much work for all except for you. Yeah, so I think that's something, not just because I don't wanna do all that work, but I think in order to make it possible for various people to do bursaries, I think it's not healthy if bursaries is one person and also there might be other things that I could be doing. So we may think about how to make me not be a blocker in the process. And one thing that's worked reasonably well during GEPGONF is that Nigel and I have been in tandem working through the reimbursement process. So I think two or three people could easily still keep coherence, not have any miscommunications, but also spread below the bit in this sort of processing incoming requests, formulating questions to the bursary committee, giving a little background saying, all right, here's where we stand. What should we do next? So that was a way to make it a more... I understand that a lot of the benefits from the static list that we got was came out of that Python script that you wrote. So you've already started to replace yourself with a small script, right? I think we should take this one step further because I thought that the CSV file sending back and forth thing was kind of painful and must have been really painful for you to merge all this stuff. No, it was all easy. Okay, but if you had decided to take a hike or spend the rest of your life on the tree, then we would have had the problem to consolidate or whatever, pick up your workflow. We should put this in or close by our conference management system somehow. It shouldn't be too hard considering how the process is mostly automated. And I think it would be even easier if I could just log on to the website and quickly, while I'm on a train for 10 minutes, like two or three people and check it off rather than save local files and so on. So I have a strong bias for offline operation, but I don't think that that should necessarily guide the processes. So whatever works for the team as a whole is fine for me. So there in fact, there is code to do the merging that you think was hard. So there's already scripting for that. I mean, yes, it could live in the conference management system if the conference management system wasn't replaced every year. So I personally am not willing to invest in the DevConf conference managing system, management system until DevConf uses the same one two years in a row. Oh, you mean summit? Pentabase? No, good point. So that's just my, I can't invest time in something if it's throwaway code. I wasn't trying to get you to commit to it, more figuring out whether it would be helpful. I think that in the long run, I mean, we might see this offline version as a kind of prototype and we might integrate something into the conference management system. One thing that's a little bit tricky and maybe I'm oversensitive is confidentiality issues. So the bursaries ratings in previous iterations were available to everybody who was admin on the conference website. And you have to decide if that's okay or not. And the problem with Penta, some people must still remember Penta, I know Martin does, is that these administrative privileges were never revoked. So the list of administrators strictly grew over time. And so eventually it became more or less 100% that if you or your closest friend were applying for bursaries, you could browse the team's ratings. Okay, I mean, if people in general can live with, on the one hand, it's clearly not ideal, right? And on the other hand, we always have this issue with system administrators and we somehow trust Luca, so I don't know. No, I mean. He's from Canada. Yeah, so. It's meant to change in the future. So I don't know, that's something to think, I would say that's something to think about in terms of integration with the conference management system is there is some advantages to being your own little kingdom, which is, but there's this trade-off too. There is an API and so you could implement it as a microservice outside. There's probably technical solutions to this social problem, yes. I have another question. In the past, we've had very few cases, but we have had cases where people requested sponsorship and then didn't show up to the conference, or otherwise like behaved in ways that were not, you know, I mean, obviously, if you get the measles two days beforehand and then we're glad that you don't show up, but if you just don't show up, which there was at least one case last year. Do we have any processes in place to handle that? We don't. The first thing that I started tracking last year, and please, I hope I remember to do again this year is Canvas people on volunteers, how effective, how helpful they were. And some members of the committee told me that that information was quite helpful in terms of evaluating this year's nurseries, but that's only taking the positive, you know, I'm a positive guy, but yeah, I don't know. There are all kinds of issues to think about once we start tracking these things. If we have, I mean, as a sort of organization of open and good people, if we have secret blacklists, it's not that we can't do it, but there's complications. So yes, we need some way of having a collective memory that doesn't get us into legal trouble. Just, you know, German law is crazy or is very protective of privacy. So we can't keep it on a server in Germany or something? I mean, I don't know. No, I'm, so more seriously, we need some way of some process to, and I don't know what that process should be exactly. I don't know if blacklists is the way to go and I don't even know if that is the reason why I asked the question, because what I really would like to have is not that we can tell the person, like, ha ha, you didn't come last year, so we're not going to sponsor you. I would like to get those people who apply for sponsorship to be more diligent about their application, so that at some point in time, in the registration process, we actually get more reliable numbers than we do in the current process. Because currently still, I think a lot of people just sign up, do the whole sponsorship thing, and then, you know, like, if something inconvenience happens or they didn't actually mean to attend in the first place, they just canceled relatively late in the stage, more work for us, less money available to other people, and so on and so forth. So if it became clear that, you know, if by the end of March you're not committed, or you'd like pull back later on, then this is possibly going to negatively weigh into your future evaluations. So in fact, some people who canceled this year said, ask that question, will this negatively reflect? I think we'd be in a better position to take some kind of moral high ground position if we managed to get the initial offers out earlier. So in order to tell people you are bad for not cancelling at the last minute, there needs to be a distinction, a real distinction between finding out and cancelling, but I mean, we could have some kind of confirmation process if we, for bursaries, right? I mean, we could minimally ask people, let us know if, so currently we say, let us know if you're not going to accept this bursary. And so that has the failure mode of ending up in people's spam folder and assuming that they're coming when they're not. So the other failure mode is people, we cancel people sort of when they really intended to come, but just didn't manage to deal with it now. Should I, it's me? Fine, I'll stop talking. So we only have five minutes left, so there are other, Medi wants to. Okay. So you added this question about requirements for depth camp funding. Yeah. Is this this? Yeah, so this is something that- How about your plans on this question? Right, so this is, didn't work so well this year, in my opinion, because first of all, I think there wasn't a clear under collective understanding on the bursary's committee of what are our requirements for granting depth camp sponsorship? And that meant, and so I personally thought, and when I wrote up the instructions for people applying, which many experienced DD's manifestly did not read, I said, if you're applying for depth camp sponsorship, please make sure you, in this text box, tell us your depth camp plan. Where the text box was somewhere else? So yes, you had to sort of switch back and forth, and that's an issue with the form. Basically, there was a question of how can we construct a process that requires zero hours of web dev, and so that was pointing people through some instructions to read. So I guess my idea was, well, if you didn't bother telling us anything, we'll just reject your depth camp sponsorship, but the committee told me I was being too strict. So that's fine. I mean, that's why there's a committee, and not just me, right? But I guess, I don't know, let me take one minute of our remaining time to articulate why I thought this was even reasonable to consider denying depth camp sponsorship to people that honestly, we're pretty confident we'll do something reasonable in depth camp. So offhand, it sounds a little bureaucratic and insane. The reason is I wanted to try and create a level playing field between people new to the community and people who everybody knows. So the people who everybody knows, if sort of get a free pass here, I guess we can decide as a community that that's okay, but it does sort of, I don't know, at the time it seemed like it put the people who were unknown at a disadvantage, just by putting the other people at a relative advantage. If you have still time, I have two questions actually. So you've seen the case of someone accepted on the schedule with the talk, but not being sponsored. Yes. Is that something that we should work on in the future, or is that an issue? So first of all, the person in question should have been better at filling out their application form and at least said that they applied to give a talk, then we would have contacted the content team for sure. If they'd said, I've applied to give a talk, we'd say, hey, should we fund this guy? He says he's, if you tell us to fund him, we'll fund him. I mean, that was my position with speakers this year, was if the content team wanted somebody, then that I considered a pretty strong endorsement. I mean. I think his point was about the schedule or some talks were published before the sponsoring were... Yes, so there was a potential communication problem, for sure, like we didn't know who was sponsored or who was accepted for talks or not, but I'm not sure the content team knew either who was accepted. So Allison is shaking her head and saying, that the information just wasn't decided yet. So I don't know what could be improved there, except... Maybe work on those. I mean, I guess in the no mail, there's a line which says, if you think there's some issue here, please contact us. And I think rather than, I guess, although it's a little less transparent, I think it would be better to take this up with bursaries than to make a public issue out of it, only because I think it looks kind of bad if people are not funded, they make a big scene publicly and they're funded, there's not a precedent we want to encourage. So this is something we've actually talked about in terms of website tooling. If you could get a simple cross-reference list of people who applied for bursaries and submitted talks, that would help enormously. Because you did actually, for the people who said they submitted a talk, you did email us and ask, are these people going to be accepted? And even though we hadn't finalized the process yet, we knew those were good enough. So we looked and we're like, yes, we're going to accept these and we let you know. So where we had the information, I think we did the right thing, but it's a matter of surfacing the right information. And that's a pretty simple cross-reference because everybody's using the same either Alioth account or Debian account. So it's possible. Internal to the content, conference management system, I mean, it knows all these things. And as Martin suggested, if the bursary process was a little more integrated with that, then, or I just asked for the right fields in the CSV dump, right? I mean, part of that was me not thinking about this. I don't know how technically feasible that would have been, but I didn't ask for that information as part of the information that I got from someone. So we can improve it technically and not just socially. My second question was, it wasn't clear for me who was handling the diversity and GSOC applications with respect to bursaries because there is some overlap. So it was pretty unclear to everyone involved, I think. So part of the issue was for whatever, well, this GSOC thing was a surprise for everyone and then for what, I guess to decrease stress on the DevCon team, this was originally going to be handled outside the DevCon budget, right? So that was supposed to be invisible to me, to our team. Right, yes. Right, so it's complicated, right? And this is, we're out of time. Okay, so, well, you know, the hallway track continues. So thanks everyone. This was a good version of this ball. There was lots of interesting discussion and thanks for turning up.