 So, welcome everybody to the DPL wannabe buff. It's on. Our facilitator, moderator here is Stefano Zachiroli, the DPL. Okay, thanks. Thanks, Daniel. So, I started with an apology. I completely misread the schedule of this day. So, that means that I didn't expect this buff to be at 10 a.m. So, it's way less prepared than I wanted it to be. But, okay, we could do it anyway. And the idea of the buff was to actually... Well, try to dispel some myth about the job of the DPL. I think we know in general in the project way less about what the DPL do than what we should know. And also to ask you for a review of... Okay. Hi, everybody. So, the idea was to share some experiences among past DPLs, and now I can use the plural, given that Steve has arrived, and people who are interested potentially even in becoming DPLs explaining a bit what their role is about, and also ask you for sort of a review of the... especially of the communication stuff I've been doing and how I've been doing that to see what we can improve. So, I guess the first message on what the job of the DPL is about, which I want to share, is that it's way less about changing things than how it is about doing routine stuff in the way you see fit. So, I remember having read a lot of BDL. Do you want to come on stage? Steve, you too. You know, that's an advantage of being DPL. You can call the whole DPL on stage and it will come. Yeah. So, yeah, the first thing which surprised me a bit when I became the first few months of being DPL is that I had an idea proposing my platform to change a lot of stuff, but actually the actual amount of work you need to put in day-to-day activity is way more than you can expect. So, you will receive a lot of requests to moderate conflicts, press inquiries, to interviews, and a lot of this routine work, which is actually a lot of work, and in essence can reduce a lot the time you have to actually change things. So, my first suggestion is that if you're planning to run for DPL, well, you need to have a good workflow to deal with day-to-day activities before you think about what you want to really change in that way. So, just a quick poll. Who in this room have ever thought, even for a moment, and then forget about it to run for DPL? That's good enough. I thought I would have seen no ends at all. At least one of them has been accounted for. Yes. So... You should have a good election this year. You should have a good election this year. So, Ian, we've run out of chair. No, come on! So, I think I want to turn the rest of the discussion with you and I welcome input on what do you want to know from people with the experience of being DPL in case you want to run for DPLs? So, we are here to answer your questions. But also in particular, I want you to think about the relationship among the DPL and the project and think about the kind of communication I've been doing and we have been doing, if it has been transparent enough or not, and what do you expect more from DPL than what you have actually got? So... We're open up for questions, comments on this stuff and all this. Whoever has raised the end, I'll try to keep notes. I have just a different question, but it concerns me a lot, is that I have the feeling that half of the DPLs just left Debian. So, we have Richard Ackermann, we have Brandon, yes, exactly, and maybe some others are not visible anymore. It is so exhausting that you can't be it anymore. Doesn't the stage answer the question? I'm sorry, you are not the right person to ask. Do we have it anymore? We need a mic for here, or we can pass this one. We have only one? Okay, I think we can... I mean, it is very visible that some of the previous DPLs have left the project. I don't actually think it's any more a higher proportion of DPLs than any other developers necessarily, it's just you possibly notice it more. I don't know, I haven't actually worked out the numbers, but you can see there are a number of previous DPLs still here. We still, me, BDEL, Ian, and TBM are all here. Me? Exactly, previous DPLs. We're all still here, we're all involved. Hell yes, after doing the DPL job for a while, there is a big potential for burnout. You don't get much rest at all, and by the time I got to the end of my time, I really did need a break, so I did slack off for a while. For some of us, we're so involved in Debian, we're not going away any time. It's really interesting, the things that happen to you the morning after you're elected as DPL, and I had an interesting conversation actually with Brandon Robinson right after he was elected DPL, and part of the reason was that I was about to give a talk about Debian at Linux Conference Australia, and since he wasn't able to be there, they expected me to be able to report a few words from the new DPL, and so he and I had a conversation about this, and his sentiment was exactly the same one I had had immediately after becoming DPL, and it was this incredible sense of responsibility. And the reason for that is that I think all of us who have been involved in Debian for a long time, and even those of you who have not been involved for a really long time, understand just how significant this project is in the grand scheme of the free software world, and to all of a sudden find yourself as the person who's expected to be the public representative for the project, the first point of contact for people outside of the project who are interested in finding out more about what's happening, the person that everyone looks to to resolve the conflicts, particularly the ones that are not technical and therefore are the most challenging to cope with, the emotional personality and whatever things. By the time you're actually elected, you sort of look around and go, oh wow, what did I get myself in for? And it does, I think, change your perspective just a little bit on what's important. But I certainly, I was involved in Debian long before I was elected DPL, and here I am long after, still hard to get rid of. So I think, to me at least, they're separate. There's no direct relationship between spending time as DPL and I certainly, as Steve said, I've had points in time where my level of engagement has ebbed and flowed. There was particularly a period for a couple of years when I was off very focused on trying to get an amateur satellite project finished where my attention to anything Debian related was probably an all time low. And I don't think any of that's unusual, surprising or to be of concern. Just a comment on that, like a tip or whatever. I decided that if I would have become DPL I would have stopped doing all the technical work for the period that was going to be DPL and that has been a very wise choice. I think we have all done that. But making it this, well, okay. For me it was really clear that, okay from this time on I stopped doing the technical stuff and it's been a very wise idea. It's difficult because first of all you cannot in just one click orphan all your packages. So if you look at my DDPO page I've listed still in a lot of packages and uploaders even though I've made clear to the other commentators that I wouldn't have done any technical work in the meantime. And it's difficult also because in a volunteer project people always try to get you engaged in stuff. So if you've been working on a package for five, ten years as I've been doing even if you made clear that you didn't want to do that for one year or two years or whatever it is people will ask you can you please do this and do that and it's not always easy to say no. This is I think a point where there is a very good risk of burnout. We are all volunteers so when a volunteer asks you something that you know how to do saying no is pretty difficult. But trying to do that on the non DPL stuff it's been very helpful for me. Penny. Yeah, I think we can. So someone asked on Monday I think whether the role was eventually going to continue being sustainable. And so my question is do you think inevitably it's going to be some house split or some particular duties are floated to other roles so essentially split and if so how? As far back as maybe was a debcon from Helsinki I think I led a discussion there about this notion that maybe the job had become too big for one person and we ought to consider even the possibility of some sort of a constitutional change to going to some kind of a ruling board or committee kind of structure or something and then there have been experiments with that. I mean Steve was sort of on both the leading and following side of the idea of a second in charge kind of mechanism to try and share the workload some. I think Zach's actually demonstrated what all of us probably knew which is that the DPL has always had the power to delegate things and if you are capable of making that sort of mental leap to I don't have to do this all myself there are lots of other people around here who want to do things and I just have to pick a few victims I mean volunteers to assign things to it's unclear to me it remains unclear to me whether there's actually a better structural approach than the one we have or whether what we just need are people as DPL candidates who use the word we a lot more than the word I. So I think there are these two moments in which you can think about splitting tasks one is going up front and saying in a platform you know I would like to work with this, this, this and this person and this is something which I don't think is particularly fair because after all the institutions put their responsibility on one person but what I would like us to try out is actually to have some sort of committee or board of directors or stuff like this which is independent from the DPL elections but is there to actually help out the DPL so if you look at the governance of more young free software project that exists right now you see most of them have some sort of board of directors and those boards actually have stuff like periodic public meetings like once per month where they review what is in need of being done and they go through it they welcome questions from people and all this kind of stuff so this is something I think we should try but it's not clear how to go there so ideally it should be something like the DPL send a call for volunteers which are people interested in helping out the DPL tasks these people show up, get delegated in some way maybe we give them some sort of governance which election, internal elections, stuff like that and those are independent from being the DPL they're just there to help the DPL and I think it could be also a good place where to learn more about the task which usually the DPL deals with and also to be more transparent because if you're up to every single month a meeting, a public meeting well it can be a way to be more transparent about what is going on than if there is only one person doing the job I think that the main reason why previous attempts of DPL teams failed is that they were mainly about let's do a team and then we figure out how it works and I think it's something that could work well I'm wondering whether something like government could work where the DPL would delegate some specific topics to someone else for a short period of time for example the DPL could decide that it doesn't care about following a specific press release that would take a lot of work that would take a lot of time to work on and delegate that to someone else we'll work on that and report to the DPL and do you think that this could work and how much of your work could be removed by doing that kind of thing? Okay so what I'm a bit scared about when I do delegation is that they tend to be more sticky than what you initially imagined so doing a delegation is kind of I was not thinking of delegations as in the constitution you don't need to use delegations as in the constitution to do that you can just say someone to the people that will interact with him that you trust him to do that to work on that for the next two months and that he will follow whatever he decides Yes so you're breaking the areas of all this I think to a certain extent that's already been happening I don't know how visible that is I mean I was trying to get more people involved and taking charge of individual tasks like that when I was DPL I think BDEL was doing it without doing that we'd have DPLs burning out every year without fail so yeah absolutely it's the better way to go I think Can you give examples of things that were delegated that way? Well there have been plenty of delegation going on so I think part of the problem is that actually calling for people on specific tasks requires quite a lot of job on the DPL side so in the immediate I mean in the near future you decide to do that it takes a lot of work to actually find the people decide the breakdown in tasks and all this kind of stuff so that's why I was trying to see if we can have sort of board which is you know autonomous which is created some way and there's some way of renewing himself which is exactly what other people are doing with boards and delegating various stuff I mean auditing stuff and I'm about to try to delegate away the spring stuff I've got some wonderful volunteers so it is going on but as long as it's there is the DPL which is the bottleneck every time finding a task and delegating I don't think it's really scalable and also I think me and you discussed sort of a government DPL gets elected and create a sort of government and then it works with the minister or whatever but one year period is actually very short and the time it takes to create this stuff it's a good deal of the one year term we appreciate if people can take notes in this DC 11 DPL wannabe goby document and maybe if I can just add what you said about those informal delegations I mean certainly when I was DPL I asked people for favors constantly and I knew what people were good at if I needed something to proofread I would go to those people if something else needed to be done I would know someone else who I could ask so you just don't see those things because we don't talk about it it's like a small thing can you proofread that press release that small thing but I mean everyone who knows me I'm sure you have been asked by me for some favor over the years that goes on all the time Just a return of experience I've been for instance asking a bit more heavily in the second term if people are interested in helping out with DPL tasks and want to sort of learn the job come to me let me know and I will be happy to show sort of to-do list and outsource to use some tasks two people have applied to that and I've shared a specific task but in the end nothing has happened yet so it's kind of difficult to do that and I think it's a problem because we have people that decide to run for DPL last minute without having an idea of what is going on and it's not particularly good for the project I think So I don't know if you were imagining like every DPL bringing in their own team a bit like say the American government I don't think that would really work for us for two reasons one is the one year thing that's been mentioned but also in Debian we tend to create roles and then people stick with them often for much longer periods of time and that's actually healthy because it means you get people who have some idea what they're doing that does mean that the DPL doesn't really get to they tend to inherit all the people who are doing everything from before and I think that's probably healthy as well because it means the DPL doesn't have to that's one more thing that's not on their plate Almost like a civil service Well there's this other element of it that I think in the medical profession is that you will start with do no harm you're supposed to try and do things that make things better and not make them worse and one of the things you realize very quickly you know if you think about this you don't even have to be the DPL to realize this but once you're the DPL it's absolutely abundantly clear to you is there a lot of people who are volunteering to do stuff that isn't terribly visible to the public and yet is absolutely vital to the project and the very last thing you want to do is demoralize or demotivate those people because in the same sense as civil service hopefully with a stronger sense of reward for whatever everyone doing these jobs is a volunteer and I think on some level we have to understand that we have to honor the contributions that they're making that doesn't mean don't fix things when they're broken but it does mean be careful that you understand what's actually broken before or taking some kind of precipitous action I think there's one question there well actually there's three questions already from IRC the first one is one moment please you say we have volunteers for the DPL election next year already the first question was that was the email address always leader at dbn.org because it sounds a little bit military like and I guess the BD knows that you know there's a point in history where the new maintainer process was waving and smiling at the right person there was a point in history where we did uploads to the archive without using cryptographic signatures on the uploads where there was a single login to the master FTP server that had a shared password you know history is full of all sorts of things like that there's certainly no military intent here you have to pick some token and in this case our constitution refers to the elected person as the Debian project leader so the word leader comes entirely from that and not from you know I think you know in hindsight calling this person the leader has had some interesting connotations in that regard I don't know what else you would call this role but it's yeah so I'm responsible for this at the time at the time when I when I wrote all this down we just had Bruce disappearing off and we really did have a leadership role that was somebody who thought they could tell everybody what to do I mean it didn't occur to me to call the job something else there is an advantage and having called the job the leader you know has to be something out of Debian or the leader is the obvious thing to call it the leadership role in Debian is really very weak compared to most leadership type roles because the authority is fairly limited but on the other hand you do have an enormous amount of credibility which is useful both within and without the project you're really the only person within Debian who's got a broad elected mandate and so that means you can stand up and you can say well all of these people voted for me so you know in some sense I represent them all and you should do what I think you should do because everybody else thinks so too effectively and the other thing that's really useful about this is a term is that it gives some handle for the rest of the world to talk to us and one of the I mean we're not talked about that very much officially in the list of jobs but anybody who wants to talk to Debian really you know if you're some journalist from a trade rag you've got really no idea what's going on and having somebody who's called the leader who's got a fixed email address gives you somebody to email and also it means that you're pretty sure that the replies you get from that person are you know in some sense representative of the project and not some kind of Debian of course being a place full of wackos Yeah the next one is well it's actually a question to all of you up there what was the most unexpected task you did as a project leader I can tell you the most of the tasks I got to do were actually expected I can tell you the most unexpected mail I've ever got as DPL like one week after my first election and there was someone proposing that Debian drop support for the Linux kernel and switch to MAC MACH kernel and getting away of all the other kernels because they are crap compared to MAC maybe others more serious replies answer to the question you know it's been a number of years now and trying to remember details is not always easy but I think the thing that I was the most I generally was how much interest there was in having me go to various places around the world to talk about Debian and I eventually realized it's the because my immediate predecessor didn't like airplanes and wasn't interested in traveling long distances so and then this was also sort of the flames were fanned a little bit by the fact that I had good support from my employer for travel expenses at the time I don't think I had any idea when I signed up to be DPL that I would spend you know the better part of the year traveling to strange corners of the globe to give talks about Debian I think every DPL since then has had the opportunity to go to some really interesting places and talk about Debian and I don't know why I was so surprised by that but I really was I didn't expect to spend as much of my time out sort of being a public figurehead and a representative for the project as people seem to want me to be at that point in time the thing that actually probably shouldn't have been a surprise but did really hit home was when I heard the news about Timo Soifer being killed in a car crash a few years ago he was a good friend and obviously just on just on that level it hit home but also to be asked by journalists and people outside the project essentially to lead the tributes to him to you know like they were asking how did he fit into the project what had happened you know how would we be hit by his loss I said it shouldn't have been a surprise but it was a very humbling thing to be asked about Yeah I was really surprised to discover how much external interest in Debian there was from people who didn't understand it at all and to find myself the contact point for all of those people and that took up the majority of my time and I found that quite difficult actually I can't remember anything like specific at the moment but I think like as a general theme one thing I guess I found unexpected I mean I was expecting a lot of work but still something you don't see we have had people complaining over the years that the DPR is not transparent enough but just how much stuff is going on just to keep things running and you don't see it because there are people who make sure that things don't break but just on a day to day basis how much effort you have to put in just to keep things running I guess that was unexpected part of a serious answer to your question so the amount of legal or legal like things that are actually going on and to be honest that's also an example of a thing that not necessarily should be done by the DPR itself we can imagine having a team of people interested in this kind of stuff and asking the DPR when they want a decision but having done all the underlying research work so this is an example of a task that can be probably applied to some other team or the like sorry no not exactly and there was a third one right yeah but I will ask this later because I don't yet understand the question either okay what else whistle Felipe about training the new DPRs one thing that may work at least would be a fix while transitioning for a board or something like that is having the elections a little bit earlier and then having two DPRs working together for a few weeks two weeks that do you think that might work well even let's assume that it works out you can have it even without changing the period of the election you just say that for the first week or whatever they are both on board to be honest even less formally than that I've yet to find any of the DPRs who just vanish and won't talk to you whether it's before the election or after the election so if a new wannabe DPR does want help I'm sure there's at least five people here who will happily give advice probably more than you need more than you want but you know if you ask for it you'll get the advice I think it's also a fallacy to believe that somehow you flip a switch and all of a sudden a big change happens Martina we're talking the other night I guess over drinks or something about the fact that when he took over after me I had already made commitments to travel to various places to speak about Debian different conferences for maybe six or eight months into the future you have to because you have to plan these things and the conference organizers want a name and a photo to put on the program and so forth even just on that level these tend to be soft transitions but I can't remember any transition between DPLs and the history of the project really that didn't involve a lot of overlap a lot of communication a lot of advice and guidance and I certainly personally feel really privileged to have been asked questions by and had the opportunity to provide some advice and guidance at one time and other to everyone who's ever been project leader all the way back to Ian Murdock even so and then I've got an anecdote from Deb Conn three years ago I remember getting into an elevator at the conference and meeting up with Margava going down to breakfast where she told me that morning she'd had a nightmare and she woke up believing for a moment she was the DPL and it was hilarious it was a good joke on the way down to breakfast and she said at that point no no no she would never do it and yet in 2010 when she did stand in the election I know she spoke to me I believe she spoke to BDL I'd be amazed if she didn't speak to I know she spoke to lots of other people asking for advice and I think we were all very supportive gave all the advice she needed we got a great DPL we could have had a different great DPL and I hope she'll stand again obviously we will happily support anybody else who wants to volunteer for a really really big job and I guess the other thing to say is that if the people who run for DPL are really well connected within the project anyway because otherwise you would never get elected you have to talk to teams you have to know a lot of people otherwise you don't have a chance being elected anyway so you have good connections anyway and most people we get along pretty well so I think it would be nice to see some more documentation and some more written down some of the experience to be written down but I think the informal way we have at the moment works pretty well so some of you mentioned that it took you quite a while to get set up in your new role do you think Debian as a whole would benefit from maybe electing the leader for two years instead of only one? So there is pro I think it would but it's also quite risky so I think we have briefly discussed that starting from an input by AJ in the sorry in the past DPL elections and so the point is that yes so there is a set up time also there is some commitment you make like as Bidale has mentioned you make commitment with people and you say okay I'm gonna talk there and also it takes a while to you know get to know people outside Debian you need to interface with can be GNU, can be FSF, can be GNOME can be whatever big project which want to be in touch with Debian in some way and at that level it really gets personal so it's not that they talk to you just because it's your role they go along with you and they will not go along with someone else so yes there will be an advantage but it's also risky because if a DPL goes missing in action like six months of the election then you have one year and a half in which he might be not or he might not be very active so what we have mentioned is giving some easy exit way at middle ground or something like people can call for an election or you can stand down some easy way I don't know. I think we did at some point shorten the time periods involved with the actual election process. I remember a long long time ago we started with I think three week chunks or something so the entire process ran sort of nine weeks start to finish and if you think about it that's a fairly significant fraction of a year particularly you know because that's a period of time during which people who are candidates for DPL I think think even more than usual about the things that they're going to say publicly who they might upset with various positions or actions and whether it's really what you would like or not the time period of the election itself often ends up being almost a quiet period for the DPL where things don't happen as publicly visibly as you might hope so I think one of the steps in the right direction was doing that I've always had sort of mixed feelings about the timing thing I certainly felt at the end of the year that I wasn't done and it would have been nice to have had more time to do more things not to take anything away from Martin because I think he did an excellent job too but the flip side of that is you still haven't gotten rid of me well yeah but I was just going to say you know you're never done so there's always something to do yeah so I didn't have the experience of standing for re-election because I basically decided not to yeah exactly certainly some of our past DPLs have made that look very easy and what I want to know really I suppose is whether it actually is easy if it is easy then we don't really have a problem there you could just stand for re-election well yes you have a point but still even when stand for re-election the campaign period is some quite some work as well so you know preparing the platform answering questions this year and still even this year it took me some time anyhow because you know and then people ask for interviews and all this kind of stuff so you're right it's not such a big deal but it can be somehow more efficient I guess yeah the third question from IRC I just read it as it is so this is my first virtual depth conf I followed a lot of talks listening and reflecting shows me a big focus on developers in the sense of contributing code my personal experiences are coders really bad at marketing even marketing their own software management skills and so on yesterday I listened to the financial management accounting talk is Debian aware that for those areas you need people with special skill in management accounting and so on and does the DPL report regularly on his done work is there a handover of the business companies from DPL to the next one so that part the answer is yes those details do get handed over and I think we've taken various steps over time to try and do a better job of accounting it's also true that the actual money that is held on behalf of Debian is not actually held by Debian it's held by organizations like software in the public interest about which sometime tomorrow I think we have a BOF session on behalf of the project and one of the things I wanted to say though is that a good friend of mine who is a machinist talking in a news group about home shop machining a few years ago made the observation that the problem is that in any organization you have to make do with the people who are there and what that means is that while as a project Debian understands that there are different skills and different roles and lots of different people with different interests coming together is what makes this thing so powerful at the end of the day the people who actually are willing to do things for Debian are the ones who are available to do the various jobs and about the specific point of documenting what's going on so I've been trying to do that with the sort of daily bits of what's going on which helps me preparing monthly report and I hope will also be a good source of information for future DPLs and then there is the sort of the mail archive which helps a bit but in my experience it's not something you use you have a look at the beginning to see if there are some to do item left or something but it's not that useful I just wanted to get back to that question from IRC and also the discussion yesterday on the finances I think what we have to realize is that we are not a commercial entity out there that has to form with other commercial entities so we might choose to do financial statements but we don't actually have to do them professionally and if we screw up we are not going to lose out on it so this creates a project which on the one hand still means that we are actually focused mainly on creating the best operating system but also gives everyone a stage to do whatever they want to do and whether that's accounting marketing or I mean I'm looking at Holger over there for instance I don't think Holger was brought up to be a video professional and I think he is you know and that's because of we have five minutes left I was just going to comment I mean I think one thing that people don't realize is that a lot of people in Debian are actually not programmers if you look at what people have studied I mean sure a lot of them have studied computer science but not everyone has and I agree we need more skills I've always been someone who has always told people don't package we have thousands of people doing packaging you have to do something else as well and for example in my Debian career I'm not famous for packaging I'm famous for other things so you can definitely be successful at those things and I think we also have in the past made the mistake of electing people as DPL because there were good packages and that didn't turn out well and I think we did learn from that and I think if you look at the recent DPLs most of them are good at coordination and that kind of work but yeah we definitely need more people doing other things I'd have a quick question to the former DPLs does any one of you have any plans maybe in the distant future to run for DPL ship again that is not a fair question maybe you'd have to ask my fiance whether or not I'm allowed to for start so I certainly don't have any plans too but that depends on like any situation in Debian you go and do work where it seems the work needs to be done and at the moment I don't think the role of DPL needs my input yeah absolutely 110% agreed on that I mean like a lot of jobs in Debian you don't do it because you want the fame or anything like that you do it because you know the job needs doing it's awesome to be able to finish the job or not finish the job but do your part of it and hand over to another safe pair of hands so I would like the answer from Bidel and TBM they are both thinking about running again next year and we go ahead with Andrea's answer yeah I have a question do you think that number of candidates somehow is reflecting the problems we have inside Debian I remember one election with 10 candidates and I remember one election with only one candidate well I think it definitely I think there is definitely a reason for that I mean I think in the past when people were unhappy you know we have different candidates and sometimes candidates who promise that they would radically change things and I think you know in recent years we still have a lot of things to improve but you know people are really happy and you know people are happy with what Suck was doing so why would anyone else run I mean if things don't if it's not broken don't fix it I think it's also worth pointing out that the fundamental themes that have been part of the various platforms for DPL candidates over the years have all sort of coalesced and converged if you look even not just this year but if you look back a couple of years there are themes that come up over and over again I'll be more open about communicating I'll do more things you know publicly I'll delegate more work these are all excellent themes but it's really been a long time since we had any feel or need for any sort of fundamental change of direction and in this sense I would be surprised if there were lots and lots of people who thought that you know there was a desperate need to go be the person who tries to lead or something through that process and I think as we've all come to understand over time the role of the DPL is not necessarily one that conveys some huge amount of power to make abrupt changes anyway Okay so time is up I just want to do you have a question I just want to apologize again for not having organized this both too much but we provide just a good show for pictures I guess then I want to observe that we suck at taking minutes and we didn't get at all to the procedures of how the DPL communicates with the project if you have feedback about that please mail me or catch me in the corridor I'm always looking in ways to improve that part Alright so let's thank the speakers coming up next is going to be talk about the NM process by Enrico Zini Thank you guys