 Rwy'n meddwl, wrth i gael ddechrau. Fe yna'n meddwl'n meddwl, dyna'n meddwl, ddim gallu'n meddwl i'n meddwl. Mae'n meddwl i'r meddwl i'ch meddwl. Felly gallwch bod yn meddwl yn meddwl, Chris Led. Mae'r meddwl. Felly ddyn nhw, rwy'n meddwl i'r meddwl. I's a bit reflect and ramble on and, you know, talk about me, the internal and the external role of the DPL and things like that. And sometimes I'd like to say something, like well, you know, has there really a power...? Makin diolp in great gain from our role. This system out there. Thank you. Roedd defnyddio eich rhaid i'r rhaid o ffordd, a rhaid i'r rhaid i'r rhaid i'r bobl, haf ydych chi'n meddwl i'n rhaid i'r bobl, a'n llygwch nhw, yma, a'r ychydig o'r q sudahr nhw fyddech chi yn gwneud nad yw'r cyffredinidol oddyn? Mae gwasgwch am yr un o gydig oherwydd mae'n digon neu rhaid i'r cyffredin, Allyn fel eu ddenyddio, oherwydd dnewol arall, o alwch chi i fyfyddo i fyfyddo. Efallai yna'n wneud. Wrth i fyfyddo, dyna hwn o'r tro refrigeratora ar gyfer gadgefodd. Mae'r dweud yn gwneud pethu o ddweud i fyfyddo i fyfyddo i fyfyddo i fyfyddo i fyfyddo i fyfyddo i fyfyddo i fyfyddo i fyfyddo i fyfyddo i'u ffordd. Rwy'n hoffi, rydyn ni'n grwyb yn ar hyn o'r lain i fyfyddo i fyfyddo arbryd gyda Siow. ..o beth sy'n fynd i'ch gwmaint sydd wedi bod eich falm, sy'n mynd i fwyaf pa'r gweithio. Roedd i chi i fynd i amgylchedd gyda'r allan, i'ch leisio, i digwydd i meddwl i'r llwyr, i'ch sioldau... ...eg mwy oedd y profiad yn eu bod yn ei wneud modd ac yn fan hyn yn credu'r lle chyhlwch mor hynny. Aeth ydyn ni'n bryd i'r ffordd amser i, ac mae'n rhaid o't oed yn eu chweithio, maen nhw'n digwydd. 5155 – I don't know if they have all recognised this. It's orange stream, CGA, things like that. I really love the keyboard. Really, really nice. My first actual Linux history was redhat. I got this from the Linux Emporio, which was a UK company that sent CDs to people before anyone has real internet. I heard the Red Hat. That's a thing. That's in magazines. I'll get that. Went to install it, though, and it said, oh, I'm sorry, you need a... I went to install Red Hat 4, or whatever it was, and 3, maybe. I got really annoyed because I'd spent, you know, £7 on this. Which was a lot for me then. I was quite young. So I was like, oh, grumble, grumble. But luckily, when I ordered this, there's a box you can tick saying, send me any old old CDs that you've got lying around. Typically old releases of stuff that they just can't sell anymore. And this happened to be a W disc. So for all you archaeologists and historians, that's potato. And I still have the CD. This photo is from a few years ago. I don't actually fold a CD drive. So when I found this, I actually went and downloaded it, and I so included it. So moving on a bit. My first Debian experience, I was using it for a few years. Things like that, changing between Katie and me. But this is my first patch over a cold December. What you don't see here is that I actually screw up the control. I still can't use the VTS properly. But this was great. I just felt great sending this off. This was an RC bug at the time in Lilihont. What really got me is that Thomas, riding back saying, rock on, we'll get this into edge. And that was just amazing. That was just like, wow, okay. I went from nothing to this is going to be in the release. I think I came out in April of the next year. And that was just so rewarding. And it was great. I went to the actual East Party in Cambridge with surgeon and things like that. It was great. But it was this sort of enthusiasm that really got me there. Thomas was not around the project anymore, but we should all do more for this. I was completely nobody then. First email for the VTS. This was a real lesson to me and should be a lesson to all of us. So, yes, these days, well, moving on from there, I became a GSoc student dead in live under Daniel Bowman. We did some cool stuff. I made a really nice interface for making live images and things like that. And this is a very fast-moving project at the time. I think we've updated the splash screen a bit. These days, at a new event, I did some ex-packaging. I added the Nuvo driver to deviant when that was born, the even shiny. I have some more random deviant-related hacks in my own, like the apt-FS file system. You mount this magic fuse file system and it pretends to be an apt repo and it unpacks all your source. The deviant-developed changes IRC was a disgusting, twisted logic line for now, someone else has to maintain that. And I did the annual wall packaging and everything I did this summer. So, yes, these days I'm doing a lot of work on the reproducible builds project, which kind of keeps me interested. It tickles my OCD, at the very least. Diffusco was a great tool to hack on. It's pretty cool. Also, I do long-term support. I'm an FEP assistant, sometimes accepting packages, sometimes protecting packages. Recently, I just got back into mid-dean hacking, so that's all pretty cool. That's just a summary of what I'm doing. Hopefully, I'm building more approachable on things like that. I mean, in my spare time, I've played a cello. I have done this in cells 4. I didn't show you the picture of me while I am 4, because I'm a bit embarrassed at the bottom of that. Okay. I also play the older gamble and the loot. So, now I'm working on the real things. So, I want to talk about what's good and what's bad, and deviant and things like that. So, what's the good in deviant? Obviously, stretch is pretty good. That's pretty obvious. And what's good about deviant? Well, deviant itself, you know, and so on. And then this CD case is pretty cool as well. We're also doing the upstream of choice for these derivatives. I mean, these are the ones I can just find logos in in about two seconds, so if yours isn't here, don't hit me too much. But yet, this is a sign that we're really sort of... People even making new distros today are based in deviant. And that really speaks to what we've built, how flexible we are, how reliable we are as a sort of social project in a sense of are we going to go in one direction one day and one in the other? No, I don't think we are. So, people can just be very reliable and hedge even their emotional interests onto us, because we're not going to just pull the rug underneath them and decide to take the operating system in different directions. You know, we've got some cool things, like I'm not sure everyone saw this, but we were on Russia today during the launch of the Orbital Inspector, which I think was last year. That's pretty cool. Well, I haven't found that. The other nice things is that we keep, with a spearhead of quite a few initiatives, so, I mean, this is slightly blown up by referencing reproducible builds here, but what other distros would have pushed this forward? Would they have had the time? Would they have had the long-term vision, the sort of that kind of thing? And I think Debbie's really uniquely placed to do these kind of things, where it's sort of a bottom-up approach, means that someone can come along one day and say, I think builds should be reproducible. I'm just going to hack on that. Let's just see how it goes. And he can go, and this shows that he can clearly go a long way. The other thing that's actually good about Debbie and this thing right now is that, although Alioth is a little bit aging, et cetera, and you may have noticed, the fact that we're planning to move away from it and we're having quite a bit more discussion about what software we're going to use and bringing out streams and things like that, I think that shows that a real healthy thing about the project right now. So although this is currently not the best thing in the world, we're moving forward and it's not reached a completely critical point just yet. We're also still having fun, so I'm not sure you can see, but I'm not going to tell you who it is, but someone every month is donating lit cents to W. And this is great. So I think this really shows how we're still got a bit of a sense of humour and things like that. We get a lot of donations and some of them are really hardening and sometimes you see the names that you can think and the amounts you just think, wow, this is great. People really care, they still care about Debbie in a really meaningful way. We still have fun, we still meet up and we're all here, so this was last year in Austria. I think we really value these meetups. It's not only the hacking we get done when we're here, it's also a method in person, it's just so much friendlier. We'll go away from this dead pump tired and there won't be many uploads in the next week, but maybe not, but we'll just feel a bit more connected to the project and to these things. That's a slightly irrational part of human beings, but I feel like it's really important and the more meetings we can arrange, the more we can go to, we'll just all feel better and the project will be a little smoother and we'll just enjoy it more. It's just so fun. Moving on a little bit more, the badge, well, you know, not everything's great at Debbie, so as I mentioned previously, I think we have two main problems in Debbie around communication and perception. In terms of communication, that's mostly around, I think, since becoming DPL, it's been about interactions between teams, primarily, rather than personal interactions, because, well, I'll come and talk about that later, but one of the biggest problems, I think, is the expectations people have of teams are often not met, and often that's because the teams themselves are just not in a position to offer something, and so if we all had better expectations that, you know, this team can do this on this time frame, if that's really long, as long as they knew when they were asking, it would just smooth everything out, maybe there's a better route, but what I'm seeing at the moment is teams being asked to do things and it kind of gets dropped, because the team has an amount of power, and everyone gets a bit of stress, and it's not really when that team is per view to begin with or things like that. So I'm not really sure the right solution in each team could have some sort of page talking about what we can provide and in what mode and how long you don't wait until you get that, but I'm not sure that formalising really fits. In terms of perception, I think we do not the greatest job talking to the quote-unquote outside world. These are some random posts on Reddit, mostly about like firmware, installation, the kind of typical things you might not find on any support forum, and we, yeah, it's sometimes quite sad really in these things, and I think we do much better job at not just making it easy for new users, having empathy with the first-time user, so when we load the Debian install or whatever, or things like that, we kind of know the kind of problems really, but we need to jump out of that and say what if we were using this at the first time, and that's just really hard in general, and I think it's just, it's even harder for hackers and things like that. I think we do impring even more on Clint Eastwood's copyrights and what's the ugly end of it, because there's, yeah, well, obviously, first thing's ugly. So, yeah, we still have some very, very bad personal communication, so sending that to a new user is just, I don't know words really, but we should be doing that in 2017 to anyone. We can't read it. Good. It's a not very welcoming response to a new comment who may have posted to the wrong list or something like that, but it's a trivial administrative error. We're just going to focus on, you know, what, how big of a wrong one. There's also a fair amount of arguments behind the scenes, unfortunately. This happens to be, well, you don't want to read this one, quite frankly. It's not even safe to work already. But, yeah, I mean, that's hate mail to myself and many, but there's also a lot of dramas behind the scenes as well. They're not necessarily ugly, but they're more tragic and it's just, yeah, these things are annoying. So what I really want to talk about now is my vision for Deviant, and hopefully it will be interesting to you or what I like about particular aspects of Deviant and how we should really crack them up. So this is like a motivational part and things like that. And what I've always thought and sort of coming from the sort of start-up world and things like that is one question that's always interesting is to ask what is a company or organisation betting on? Like, what are they kind of hauling on? We're on this idea, if it doesn't work, we'd kind of go bust. Now, that doesn't really fit Deviant, but it's just a really interesting framework and question to ask. So, I mean, you can look at organisations like the Free Software Foundation. They went all in on Free Software. They didn't try and do other things as well. So what is Deviant better than these things? Like what risks is it taking? I think we could just say a lot more risks. I mean, this is the rather infamous open SSL change. And this could be bad with someone saying yes. But this is someone taking a risk and I think we just take a lot more. It's reversible as well. Most of these changes, you can roll them out. I mean, we have the experimental suite which I think is completely underused. I mean, what is it? One, two, three percent of packages have an unstable version. We should be using that a lot more. Or just unstable, you know, things break. No one's going to hit you too hard. But I think we're growing up in a culture where we aren't jumping on people's backs as much. And the more people are less afraid of load things, they're slightly risky, just taking more chances and things like that, that becomes a muscle that grows and comes with extra use. We can start to drop our ego even more but it's difficult as well from a personal point of view because it speaks to people's insecurities if they don't want to see me making mistakes. But I think that we're at this point where we can tolerate that a bit more. Here's one good example of an entry recent email about that farmer and we should see a post like this once a week. And sometimes the ideas will be very well thought through. This one is very extreme. I just haven't been booted yet to enable the APR myself. But we should be seeing these once a week and having discussions like big proposal of changes and we should run with some of them. Maybe APR turns out to be disaster but we can just revert it. It's not a problem. The perfect is the enemy of the good etc. Here's one example of something that was reversed. It's now being closed. No jazz binary and now we're not changing it. We've changed it back or whatever the actual details are. It's fine, just reverse it. Who's saying, oh, I'm at the city of the tech committee they said one way, one year, one thing, the next. No one's saying that. Fine. Great. We've moved on. We've changed. We should do more of it. That's pretty much what I had to say, really, about... So, yeah. We should also ask them what help. So if people have issues this is to basically help the... to get over people's ego and things about trying experimental stuff. So if you do have an idea, I mean, do try something private, of course, but also do make posts, like, introduce it to the development and project so we can try all these interesting new ideas and push them in the right direction in that sense. So I didn't have anything else more to say and thought I'd leave a lot of it up to any questions people have. I think some people do. I want to thank you once again for lifting me as leader. I'll say it's just an honour to be up here and see so many friendly faces and things like that. Anyway, thank you. Hi, Chris. So I appreciate I'm probably not even the first person to ask you this yet but it is a long-standing tradition or you're going to stand again for DPL next year. Is that really a tradition? Oh, hell yes. Try to avoid the question. I haven't thought about it. Cool. Are you enjoying the job? I'm very much enjoying the job. It is quite demanding. It takes a fair amount of time and a different pile of spoons that I'm usually taking from and that's been personally quite a fulfilling experience. And very informative in teaching me about myself and growing as a person. Yeah, yeah, yeah. Don't call it. But yes, I'm very, very much enjoying it. Excellent. Hi, Chris. My name is Brandon and I have maybe some notion of the pressures you're under. Hi, everybody. But not to pull the spotlight onto myself as I want to do but rather to focus on a point you made which I wanted to underscore emphatically if I could. I think you're right to ask for the project to take a few more risks. The point I really wanted to zoom in on though is your emphasis on the under-usage of experimental versus stable. And also this point ties in with the whole issue of teams getting over work. I'll tell you firsthand, I haven't really said this in front of a large audience before, but managing the issues between teams was the single hardest aspect of the job for me. I don't think I handled it well. So if you want to say I burned out and left, if you could hang it on any one thing, it was that. Because, you know, even though I was notorious for getting in arguments with people, I really care about everybody in the project then and now. And I'm sure you're experiencing some of this yourself. It's difficult to work through some of these issues. And the key I wanted to focus here on is expectations management because that's the phrase that popped into my mind was something that you hear a lot in corporate culture, SLA, service level agreement. And you said, well, do we really want to formalize our teams that much? And I would say there's a risk in not doing that, which is that people will make assumptions anyway. People are going to accept a certain level of performance from any team or even in the individual package management. And if you don't make these explicit, then, you know, 200 different people are going to have, you know, maybe 100 to 120 different ideas differing about what can really come from that. One of the beautiful things about the Dewey and Policy project, which goes all the way back almost to the beginning, is precisely that it establishes that set of expectations. I don't expect team management to be quite so easily codified, but that is the direction that I would recommend that we explore as a project to try and address this issue. I'm not surprised, but I am sorry to hear that internal team struggles, more between team struggles, continue to be such a large, taking a large part of your time as I've seen from your bits. And that's the most wisdom I can offer you with the hindsight that I have. So thank you for taking on that role. And good luck. And the job definitely happens first, so please enjoy. In case you don't see me for another 12 years. Thanks again. If you want to speak to me after this, you're obviously very welcome. You all know my face now. So I can't be as anonymous as I was before. Anyway, thank you again, and I'll see you around the rest of the week.