 ..uncle Steve, Sledge, whatever you want to call me. Please, not right now. It's my pleasure to stand in front of you at DebConf. And, well, give the first keynote. Mum told me to think that, of course the audience here will be nice and friendly. I guess she'd never met you, folks. So. Debian, when did it start? 16th of August 1993. I'm sure we've all seen the email that Ian Murdoch sent. It was announcing to the world the brand new Linux release that he was starting work on. It was built from scratch. He'd been using SLS previously, and while it was functional, he had quite a lot of things that he didn't like about it. So in the best traditions of free software, he started again. So he was inspired by SLS, but there were lots of things he didn't like. He wanted Debian to be sleek and slim, but with lots of documentation. He wanted really easy installation. There's a whole load of other things as well in the email. He wanted it to be to always have the very, very latest versions of all the software. He didn't want to have humongous amounts of random garbage on a system. He wanted it to be a nice user experience. Of course, fundamentally at that point, he was the user. So whatever he wanted was what went in. Fifteen years ago, the world was quite a bit different for Linux, of course. Back then, there were only a very, very small number of users compared to what we've got here today. But some things were still the same. Although there may not have been a huge number of users, there were still lots and lots of people working on their own distros. Everybody had the freedom to do what they wanted. They could put together a system that just had the bits they needed. They could do their own configuration. All of that, of course. The freedom to do your own configuration never really went away, because at that point you had to do all of it yourself. Nobody was really going to make life easy. Ian's idea, of course, was to build a free distribution based on the principles of GNU. I think it worked quite well. Either that or all, very deluded. So, 15 years on, I am, I think, the 11th project leader. Starting with Ian, first parents, of course, Ian Jackson, Wicott Ackerman, Ben Collins, B Dale, who, of course, is here today. Martin Mickelmire, Brandon Robinson, Anthony Towns, Sam Hostervar, and me. We've had, again, this is actually more difficult to track. More than 10 releases. If you actually, depending on how you count some of the early ones, they were numbered. The versioning on them wasn't particularly consistent from one to the next. Does that sound familiar to anyone? And then, about Booth's time, we started using the code names that we've all come to know and love. So, we went through buzzer expo, Hamslink, Tater, Woody, Sarge, Etch, and coming soon. Lenny, yay! Yeah, we'll need to talk to the release team about that, I think. So, quick show of hands. Oh, I should have said before we start, there's absolutely no way I'm going to stand up here and spread rubbish for a whole hour. I want you lot to participate as well. So, hands up how many people were around when using Debian before we started the names. Two, three, BDEL, Phil, and person at the back who, yeah. And then, with Buzz, Rex, Bo, Ham, I was kind of hoping we'd see more and more hands going up, but fine, okay. Yeah, might not be a bad plan. Slink, Potato, Woody, Sarge, Etch. I'm hoping that that will probably cover the room. And coming real soon now, Lenny. We've had a long history of producing free software. That's something that we shouldn't ever forget. We do have a long, reliable history of doing our releases. Maybe not as fast as we'd like, but we've had thousands and thousands and millions of users who've followed us along this track. So, today, we started off Ian's initial release. It wasn't very clear what he considered as a package is or what he considered as a release. It was quite fluid. Today, anything but, I had a quick look last night and I'm scared myself. At 4386 in Lenny, we're currently over 22,000 binary packages. Ow. I'm not sure how many source packages that corresponds to, it's probably just around 10,000 at the moment. We've got over 1,000 registered developers. Now, Ganeff at the back, if he's still there, can probably tell us exactly how many we have. Yes, no? We'll come back to him. We've got a lot of developers. Now, obviously, we are aware that not all of these people are necessarily active. It is all too easy to just go quiet and we don't necessarily pick up on people, but still, we have 1,000 plus people. We are one of the biggest free software projects. We have another, I'm guessing 2,000 people from the sheer amount of activity I see on the localisation lists and new maintainer on the sponsors list and Debian Mentors. We've got even more people who want to work with us. They want to become DVDs. They want to help us produce the best free software distribution going. How cool is that? I'm guessing you've probably picked that up a lot today. A lot of these numbers I don't actually have hard figures for. Yeah, I'm crap. How many users do we have? We don't know. Honestly, because, of course, we don't do any of this rubbish like ask people to pay for what we do or buy certificates of licences or anything. We're guessing we've got millions of users. Anybody got any better ideas? Cool. Right, that's from one country. I'm sure that we'll have similar from... Yeah, we'll have similar from all over the world. We do push a staggering amount of data. And, of course, the latest thing is that I'm sure people will have spotted we're now even at the stage we've been around long enough that we've got Debian developers marrying each other. The way things are going, and I'm sure Btel may well be one of the first people to be a proud parent, to have one of their children getting into Debian maybe. No, it's... Help, we're breeding even. So, how did we get here? Why are we all here today? What's the core of Debian? Fundamentally, we've got a lot of common beliefs and ideals. We're all working together on this big free operating system. We may not necessarily agree on all the details, but we've got enough of a common core that we can work together. We've codified that common core. We've put the details in the social contract and then in the DFSG, the Debian free software guidelines. Again, I'm assuming hopefully reasonably that everybody here knows what they are, because I'm not sure it's hella not going to go through them point by point. We're not always the most friendliest people every minute of every day. We're all aware of the occasional arguments that go on. We have... Yeah, we have the occasional disagreements. We have good, strong technical discussions. Sometimes they do just evolve into flame wars. To be honest with the 1,000 people with such strong opinions, that's hardly a shock. The main reason, at least for me, the main reason that we're still here, that we're still doing this, is really good fun. Hands of anyone who's working on Debian who's not having fun. Right. Seriously, if there is anybody here who's not having fun working on Debian, please come tell me what on earth are you doing? So, what are we good at? We've got massive dedicated developer community. We've got loads of people doing lots and lots of really cool stuff. Whether that's packaging, just doing pure packaging work of upstream developers, whether that is working with upstream developers or even in a lot of cases, just being the upstream developers. We've got a lot of people doing a lot of work. We've got a huge number of packages. We are without a doubt the biggest distro just purely in terms of the number of packages we have. I said, 22,000, and I know this because this is something that I do watch every day. 24,000, yeah, okay. Is that in, well, I386 main was 22,000 odd when I looked last night. Yeah, that's not much. Oh, sorry, I forgot. Gannas had time to do the new queue. We're looking, our next release is going to be running on to probably five DVDs just for I386 binaries. Or, and I will brag, I've got blueways working. We will have some blueway images. So yes, you can go to your friends with one disk and say, here's all the free software you ever need. Now, in terms of architectures, we support more than any of the other Linux distros. We don't necessarily support everything that we once have. Apologies to the 68K folks. You didn't release an etch. It's looking very doubtful for Lenny. There's a chance that a couple of the architectures that we supported in etch may also not make it for Lenny. But even if they don't, we'll still have, I think that will be 10 architectures. Nobody else comes close. And then finally, we've got more and more teams working together. As the workloads on individual developers grow, as we end up with more and more work for people to do, people naturally fold teams. It's nothing particularly centralised. It's nothing particularly formal. But oddly, we seem to like working together and sharing a load. That's cool. Now, what's bad? I'll ask for any more in a minute. We have struggled to release a few times. Again, this is not news. The time that it took to release potato and woody and sarge got quite painful. The delays, the issues that came up at the last minute, to a certain extent, lack of planning, lack of control, they weren't good. Flame Wars, we've got 1,000 very, very motivated, very, very opinionated developers. On any given issue, that means we've possibly got 2,000 different opinions. When they come together, why, God, do we have some epic flame wars sometimes? It's not always bad. Sometimes, out of the middle of a really heated discussion, some good things happen, but not always. It will be nice possibly if people sometimes stopped and thought about what they were saying, rather than just instinctively reacting to what they see as an insult or similar. We're human beings. There's a limit. But sometimes we often don't necessarily work all that well together. One of the things that I picked up from the teams review that I did, that I promised when I was elected, is we do have quite a few cases where the various teams are not all working all that well. We have people who are essentially one man teams with other people just not contributing, or we have a couple of teams where there isn't any well communication. It's people, somebody checks something in, and the only communication they get back is somebody else who revert to the commit. That's not wonderful, and it's something that we need to work on. Can anybody think of anything else that we're not very good at? Anyone? Holger? Yeah, PR and artwork. Anyone? Wow, full of sleep. Um... There's always more for us to do, okay? In fact, the converse of that as well, sometimes we're also not very good at pushing our own changes upstream. The community is relying on people working together, and we do our best. I think we're probably one of the better groups for doing it, but we could always improve. So, why does that matter? Well, scary thing I've had since being DPL is I've had so many journalists coming to me, asking me the obvious questions that, of course, they could have picked up by just having a look at the website. Okay, that's journalists. But equally, we've had a lot of them coming to us, not just asking those questions, but asking for my opinions, or hopefully more generally the opinions of the project, on a lot of the issues that face free software today. We've been described, and I've been told this by more journalists than you might expect in just the last couple of months, that they see us as the most important distribution. Now, obviously, there's a lot of that is going to be flattery just when I talk to them. But honestly, sorry? Yeah, as I'm about to say, it is true. We do possibly the most work of any of the distros out there in terms of making all this software work together. If you have a look at a lot of the other distributions, I'm not going to name names, it's fairly common. People can put a lot of time and effort in to developing small parts of the system, or even large parts of the system, but they don't chip anywhere near as much software as we do. They don't have the same number of people working on it. They don't have the same relationships in between the software that we have to deal with on a daily basis. I mean, this is why we've got so many developers who are busy all the time. It's why we've got a release team these days. How many people in a release team at the moment? 10. And they're busy, as far as I can see, because I do watch the release list, honest, we've got 10 people working typically several hours each every day just tracking what's going on, just making sure that the next version of Debian is going to be releaseable when it's going to work. That's a lot of effort. We are upstream for actually quite a lot of the core of the Linux system these days. Stuff like System 5 in it, stuff like Goff and the MandiB stuff that I know Colin works on. There's a whole host of other software I'm not going to pretend to know all of it, but we actually do a lot of the core development, a lot of the core design that everybody else depends on. To flip that around, of course, so do most of the other distros as well. We are free software people, we can all steal each other's work. That's great. So, we've got a massive number of users as well. I said, the best estimates are it's in millions, it could be more than that, it could be less than that, we really don't know. But to be honest, if we only had a few thousand users, my God, it's important. Those people depend on us every day to help them do their job, to help them live their lives. And, lastly, possibly most importantly, everything we do is free and open. There are plenty of other alternatives if you don't want that, yet people come to us just because we're free and open, because they know that everything we do for them, they can reuse, they have the freedom to modify, they have the freedom to share with their friends. They can also get involved. If they want to make changes, if they want to help, they can talk to us. They can help us, and we'll share what we're doing with them. That was what really hooked me on free software in the first place. The fact that I could actually help, not just develop bits and pieces of a system, but I could actually get involved and help develop the operating system that I depend on every day. It's very addictive. I've lost the last slide, yay. So, I'll pretend that it's still there. Please pretend with me. We've been around here for 15 years. What are we going to do for the next 15? How are things changed? Well, as I said earlier, the Linux market back in 1993 was tiny, but a lot of the same challenges were still there. We still had a lot of people doing their own distros. We had a lot of work needed to make those distros work. Okay, 15 years on, those things are scaled up. We now have even more people working on the distros. We now have even more packages, even more work needed, but equally we've got more friends helping us do it. We've now got machines that are light years faster, bigger, more powerful than what we had 15 years ago, but I'm sure we can rise to the challenge and make them behave no faster than they were 15 years ago. We can fill those hard disks. We do regularly. We can eat that CPU. We can put nice, big, flashy graphics all over those screens and remind people that they're using W. What are the new challenges that we have? What changes do we need to make to survive for the next 15 years? We need to make sure that we continue to scale. As time goes on, we're going to have to add more and more people. There is going to be, well, based on the last 15 years, we're going to see an ever-growing amount of software coming. We're going to have even more challenges in terms of making that software work. We're going to have even more challenges in terms of just talking to each other. The more people that you have, the more time and effort that typically you spend in communicating with each other. Those are all going to be problems. Do I think we're going to have actually struggle with those? To be honest, no. We've done it before. We can carry on. What? I'll throw it up to you guys. Do you think, well, can you think of any more challenges coming in the next 15 years? Okay. And it might be that with the current approach to it we are going to face problems that we will just not be able to support the hardware which is coming out. Yep. It's something we're going to have to work on and fight against as far as we can. Definitely. Anybody else? I think this could be described as something that doesn't work very well in Debian as well. We need to be able to control our growth. We are growing very fast in terms of number of packages. You mentioned that. But we are growing from the outside. We are extending our packages' basis but we are not controlling the core of our packages. Just once look at some very important packages in Debian and look how they are maintained. You'll be scared a lot. We need to have more control on the very important stuff in our distribution to help the release happen and the things like this. And controlling this growth and maybe getting more resources. I'm not that optimistic about resources coming and coming and a number of developers growing all along. I'm not sure we will still continue to grow that way. Anybody else? The last thing I wanted to show, I'm sure people have seen this before, this is the program DevRoster, really trivial shell script. Pulls up a terminal and just sits there displaying a list of all the maintainers we have. And it wraps badly. Fine. We've got a lot of people. We work together very well considering that size. We have one of the biggest, not just free software projects, we've got one of the biggest. Yeah, sorry. We have one of the biggest projects full stop of the biggest development teams anywhere. And that includes commercial companies, commercial groups. If you go and talk to Sun or talk to Microsoft or talk to IBM, they may have a larger number of software developers in total in terms of teams actually working together on one project. We've got to number up there. One of the questions that journalists have been asking me over the last few years, both as DPL and before, is how on earth do we possibly cope? How do we not kill each other? How can we actually work together despite being spread so far across the world and still produce a usable, high quality, very large operating system? And we've got anthropologists studying us to see how we all tick. That's quite scary. I've even been asked by people at corporates to say, could I go along and give talks telling them how we do it? I'd love to if I could actually understand that completely myself. I'm just gonna leave that running for now. Right, I am woefully on the time. I will fully admit. On any questions, any points people would like to make? Neil? Yeah, I've got a question from Martin Craft on ISE, which is, why is this inevitable that we have to keep adding people? Why do we have to grow in terms of developers? That's an easy one. We always have to keep adding new developers if nothing else because older developers move away. Basically they lose the time to work on Debbie and they lose the interest. Equally, it's, I suppose, it's natural. So long as there is still more free software development going on and there are more and more people wanting to use it within Debbie, we will end up picking up people who care about it. That's fundamentally how most of us got involved, at least in my experience. We get involved because we want to, because we think it's fun, because it matters to us. So long as we can carry on making it matter to people, so long as we can carry on making it appear fun, making it possible for new people to join us, I think it will continue. I don't see any, necessarily any end to it. Any more? I have a question. The problem that I see in Debbie on ISE, there are many people that maintain many packages. And when that people maintain that many package, many packages could be lost in the time because they can release in the time that we need to use. So how can I do, or how can we do with that maintainers? There are many packages like the closer packages, like, I don't know, liveries, music, many, many, many packages, and there are few people that maintain that many packages. So how can we do with that people? Because we have to try to control that people use or make packages, or make packages like a little, little, little, little maintainer. I think you understand. You're worried that we have people working on too many packages and not spending enough time on each package. That's it, that's correct. Well, what I'm saying is that something needs to be controlled because fundamentally it's, people can choose to work on what they want. Instead, I'd rather see us encourage those people to accept help. If you can see that there are packages which are not being maintained as well as you'd like and they matter to you, to be honest, the best thing to do in the free software world is off to help. Talk to those people. If you can help them test or fix bugs or even just work out what is going on in some of these existing bugs, join them. Or if you've got friends who are using Debian who don't know anything about how to get involved, talk to them, get them interested, point them at the places where we need help. I mean, there are always, and God forbid I can be one of these myself, there can be maintainers who don't actually acknowledge early enough or often enough that they need help. Talk to them. Be nice to them, obviously, because fundamentally, they're spending their own free time doing this. Very few of us are lucky enough to be paid to work on Debian. But talk to them. Be nice. Offer to help. Don't start calling names because they haven't necessarily done the work that you want them to. Instead, offer to help them find the time to do it. I've done that with quite a few people already. It works. I might be one of the person with quite a few packages on my hand. And to some degree, I like to differ to change a point of view because I happen to stumble upon a lot of people who just maintain a single package or two or three and are not really keeping up with it. I think maintaining some more packages gets you a better working flow and gets you more on tracks with things and actively having to check up with policy changes and other things. And it reduces the time that you need for a single package if you maintain more. And I really often stumbled upon people with just one or three, two or three packages that really were in a bad shape. It can happen. Equally, yes, the cost of maintaining each package drops as you maintain more, but equally, if you only have the time to spend maybe an hour a week or whatever on Debian, if maintaining one package is what does it for you, then we need to encourage those people. Yeah, this is BDL. Unfortunately, this depends so much on what the packages are. I have found myself at various times in the past realizing that there were large, complex source packages that emitted various binary packages and so forth that I was responsible for, that were no longer things that I personally cared as much about as when I first started maintaining them. And those are excellent examples of things that you should think about getting help for or giving up to somebody else. In fact, there are some where I'm very pleased that I'm about to take myself off the uploader's list because those transitions have worked marvelously and other people are doing much better work there than I would have had time to do. At the same time, it's quite possible for there to be a situation where there are a number of things that are all very much alike and have similar sort of workflow requirements, similar kinds of relationships with upstreams. And the number of packages is frankly not really a strong driver on this. It's, to me at least, it has to do with the complexity of those source packages and the complexity of the interactions that you have with other packages and upstreams and so forth. So any more questions? Neil again? It's a Martin Crafter gang, actually. He asks about cross-distro collaboration, things like Fedora and all the various other distros will scale upwards as well. Are we ever going to work together? What opportunities do you see for greater collaboration between the various distros? First things first, can you ask Maddu why he's not actually here asking the questions himself? Oh, okay, fine, he's got an excuse. To be honest, we already work quite a bit with the other distros, especially on things like security. I know from talking to Neil and Moritz and a few of the others that we already work regularly in terms of the vendor set, less people are sharing patches, that kind of thing. In my own packages I know that there can be quite a lot of interaction between the different distros. So, again, working together in terms of the patches, working together with upstream, so that the patches that we've all got actually get integrated in a useful manner. So, hey, it makes our lives easier, we don't have so many patches in the future. The package kit folks are working on trying to share more of the work that all of the distros do. There's plenty of scope for people to get more involved and work with each other. Then, of course, there's also the high profile, we've got lots of distributions who use Debian as a base. It's actually one of the things I want to do. I'm hoping to spend more time on this, talking to those people to see how can we work together better? How can we help them be more efficient? What can we do to help them release their distributions and also share the work they do with us? So, just to make the most of the efforts that we do have. I mean, any more suggestions on what we should do? Right at the back? Yeah. I know Martin, in particular, has been talking to a lot of people. I presume he's on that list. Right. I just wanted to point out that it will always remain difficult or be impossible to effectively limit the software that will be included in our archive. I mean, it will be very difficult to effectively hinder any maintainer to upload free software that is indeed free to our archive. And I think people should, if they are worried about it, they should spend some of their time instead of maintaining packages putting into our infrastructure for handling all these packages. Because that will be the most important part that our infrastructure continues to scale or starts to scale in some cases. Because I don't see any way that we will ever decide to stop to grow. So, we should spend our effort not discussing about that, but discussing about how our infrastructure can scale and what everybody can do about that. Yeah. Of course, it's another place where we have evolved substantially over the last 15 years. When Ian first talked about Debian, he was expecting to have a single system that people then did. If people needed any more tools, they'd be responsible for compiling and installing them all themselves. We grew from there into the concept of packages, and most versions of the package written in pull was very simplistic. We moved on. That was before my time. We moved on to when we had the first C version of the package, and we had deselect, and we all loved it because we didn't know any better. Some people still love it, yes. Of course, a lot of the challenges that we had then wasn't so much managing how Debian worked. It was managing how the actual system on people's machines worked. Back in those days, when we had maybe a couple of thousand packages or even a few hundred, it was easy to let people just go through and have a look manually at what each package did. Can you imagine how long it would take to do that today? 22,000 packages? We've moved on. We've now got better and better packaging tools, better and better package management tools so that people can actually find out. We already have eight programs that do what they want. They just need to find them and install them. Obviously, we've got apps. Enrico, who I wish was here this week, has done a huge amount of work on DevTag, which is a lovely way of helping to organise the archive so people can find their software in the future. As we continue to grow, and who knows in another 15 years, we may have a million packages in the archive, but how do we work out how to find those, how to manage them? The answer is on the postcard. I'm sure people would love your help. After 15 years of work in Devian, how do you think that Ubuntu is affecting it? It's a good thing. It's a bad thing. That's an interesting question. Obviously, Mark is here in the middle of the audience as well. Ubuntu is one of a number of distributions that bases their distribution on a lot of the work that we do in Devian. They've been phenomenally successful. I don't think anyone would disagree with that, and I think that's a very good thing, both for them and for us. One of the cool things about free software is that people can borrow each other's ideas all the time. People can share their ideas, and this isn't a zero-sum game. This is not something where, if they get more users, we get fewer. The more people we have in the world using free software, being exposed to our ideas and our philosophies, the better. However that might happen, if it happens to say Red Hat or Ubuntu, or Slackware or even Gem2 on the front of the box, it's all good. All of those people out there working on free software helping to push it to users in the rest of the world, the more we all gain. I'd far rather see people with an Ubuntu box or a Slackware box than with a Windows machine on the desk fighting with a lock-in that happens every day on that kind of system. We can always improve how much we work together in the different free software distributions, but it's coming. Naturally, because we've got the same core ideals about what we're trying to do and similar philosophies, we end up meeting up, we end up drinking beer together, we end up working together. It's cool. Another one from IRC, and apologies if I can't pronounce this properly, but it's from Baz Otheku. It seems that as Devian is growing, both in the number of packages and in the number of developers, it's getting more and more conservative and it gets harder to make big changes. Do you agree with that? And if so, do you have any ideas on how to change that? Absolutely. Yes, in some cases, we do get more conservative. The bigger you get, the harder it is to make changes across the entire system. It's self-evident, to be honest. In terms of how to fix that, how to change that, we need people to help spread their enthusiasm. If people have cool ideas, I would hope that we're still open to them. There are still plenty of scope for people to come and share their ideas with us, for people to, admittedly, to defend them in sometimes maybe heated arguments on a development list, but there's still plenty of scope for new ideas. I appreciate there's a lot more to that discussion and it's probably a bit more than we're going to have time for today. I shared the worries of Olga, and I think most of us, that we are bad doing PR and more generally marketing, I think. Even if it's true that nobody can be forced to do anything in Damian, and it's rightful to say this, I think that as a project we need to find a way to drive people to do something that the project needs, and I think the example of the website is just the most famous one. Do you see any way that we can establish practices for having things that we need being done? We need to encourage people to actually think about what's needed. To be honest, the web team actually, for my team's survey, I did find that the Debian website team was one of the bigger teams where I got most responses. The weird thing about that was I'm not going to name any names because I said when I did the survey I did promise at least some confidentiality. The weird thing about that was that the vast majority of the people on the website team were complaining that we all need to get together and do stuff and stop arguing. Maybe 90% of the people who responded all said that, yet if you see what's happening on the web, actually on the Debian www mailing list, you can see that those are saying people are the ones who are not doing it. It's human nature. We actually need, in some of those areas, to actually get people to be more honest and open. Actually be prepared, not to be nasty, but to be frank about what they think the problems are. If we really do have a problem where somebody just needs to go away and spend a couple of months doing seriously hard work and then come back for review, actually do that. At the moment, again, carrying on with this example, we've had several people suggest it, but various other people on the mailing list have all said, oh, no, no, we couldn't possibly do that. That's too much effort, that could never work. Actually encouraging some people to be more adventurous, to go away and do the work and see how it goes. If we need to have more people get together in real life, that conference is great, the meetings in Extremadura are great for this. Let's get people together to meet face to face. We do probably better than anybody else in the world at collaborating entirely electronically, but no matter how good we are at that, it doesn't compare with meeting together face to face around the same table with a white border or a black border and ideally some beer on the table. Sorry, I might sound like I'm obsessed with beer, I like it. Okay, Frank, again? I think, to answer a bit directly to that, I think what people need to realize that in Debian, where you have a very loose hierarchy and a very loose association between the people on the organization side at least, that work will always be worth more than worth. A lot of the problems in Debian begin to exist when a lot of people are complaining about something, but nobody actually does anything. At many points in the past few years when something did go right, I can see that it was because the people started to work on it even though nobody encouraged them, even though nobody officially blessed them to do something, but because they did it, other people just used their work and so a good idea persisted. I think that's the same, and the same is true about PR and stuff like that. If you do something and other people see it and like it, they will support you. That's the experience of Debian. It's something I've been pushing in press interviews and whatever as well, and I'll reiterate it today. The best way of getting people to agree with what you're doing to join you and help you, fundamentally, is to go up and do it and tell people what you're doing. Be enthusiastic about it, you'll get help. Show that you're doing cool stuff, you'll get help. I think one of the things that's a little bit weird is that somehow over the years a lot of people who, even people who are full, complete Debian developers, somehow don't feel as personally empowered as I think they should, and as those of us who were around in the early days of the project most certainly did feel. If you doubt that this is the way things are supposed to work, go have a look at the Debian constitution. It's not unique in the world, but it is very interesting in that the vast majority of the power to do things is reserved to individual developers. What does that mean? You really are supposed to go fix things if they aren't right. You really are supposed to go work on the things that bother you the most. You really are supposed to lead initiatives and propose things and be willing to just go do the work in any sufficiently large organization, and we are certainly a large organization, there will always be someone who disagrees with you. There will always be someone who thinks that your proposal is the absolute worst possible idea on the planet. If you allow that vocal minority potentially of people to control the amount of enthusiasm and energy that you put into the project, things won't work. You have to be willing to have some faith in your own ideas to be willing to do some work off by yourself to show what it is that you are talking about and why people might want to care about it. If all of us just keep doing that, then we won't have a problem. Any more? Doug? Just to even amplify that just a little bit. One of the things that would be really helpful is if the rest of us who are aware that Edebion is a meritocracy, you can sit down and do that. When we see conversations like that getting bogged down in Contrarianism, that we stepped in, or not even stepped in, but send private messages of encouragement to new people who are experiencing something or have great ideas, or even just ideas that they seem impassioned about to encourage them. That's something that takes seconds to send an email, but it can change somebody's perspective on what's going on. Even how much effort they're willing to spend. If somebody is getting kudos, it will spend thousands of hours on something willingly, whereas if they're being harassed all day, it's hard for them to spend any. If everybody could do that consciously, that would be really great and help us grow even better. Definitely. One of the most important, one of the biggest advantages we have is the fact that because we're volunteers, because we've got so many developers, we potentially have a thousand people all working on what they find interesting, what they're passionate about. Compare that to even the largest company with maybe ten times that many developers who are doing a day job because it pays the bills. We've got so much more scope for people to have ideas and be creative. Let's encourage them to do that. The other chances are maybe only one in five of those ideas will actually come to fruition. But by God, let's encourage the people, even with the other four, to have fun trying it. That's where we'll get more ideas from. Anymore? I'm guessing not. I guess I'll close then. Thank you very much for coming and listening to my babble and even helping me a little. I said I wasn't planning on standing up here for an hour, as you can probably tell. Fundamentally, as Biddell points out, the constitution of Debian is a very interesting document. It was crafted to a certain extent to make sure that people like me actually have very little power. I don't want to stand here and pretend that I'm the great glorious leader deciding which direction we're pulling in. I wasn't elected for that and I sure as hell don't expect to do that. Instead, I want to hear what ideas people have got here. I want to hear what ideas people have, all the mailing lists, or wherever they come from, and then help take the best of those and improve Debian however we can. Thank you very much for coming today. I hope to see you around the rest of the week.