 I think we'll go ahead and get started. The people who wander in late are going to miss the best talk of the day. So let me begin by introducing myself and laying out a little bit of context of how we're going to spend the next 45 minutes or so. My name is Max Spievak. I work for Red Hat. I'm also a Fedora project contributor. There are certain points during my talk today where I'm going to talk to you as a Fedora person, and I will be standing as such. There are other times where I'm going to talk to you as a Red Hat employee, and I will walk over here and put my Fedora on, and then you will know that I'm talking to you like a Red Hat salesperson. I've been working at Red Hat for about four and a half years. I spent two of those years as the Fedora project leader, so I took over that role just before the release of Fedora 5 and continued in that role until just after the release of Fedora 9, almost exactly a year ago. For the past year, my job has been to manage Red Hat's global community team, so that's a team of people who participate to a solid extent still in Fedora, but also part of what the community team at Red Hat tries to do is ensure that Fedora isn't the only example of Red Hat's commitment to open source. So we work in other parts of the company as well. However, this talk today is about Fedora, and so that's what I'm going to focus on. So let's lay out a little bit of context, because I found it interesting in Joe Brockmire's open Susie speech an hour ago. He put up a slide that talked about some of the sort of large goals of the Open Susie project, and one of those goals that he said was to make the easiest use desktop for anyone in Linux. So the point I want to make here is that there's a whole bunch of people in the audience, and if I were to ask each and every one of you what you use a Linux distribution for or what you use Fedora for, I'm going to get a lot of different answers. And that's fine, because Fedora can serve a lot of different roles. Maybe you're using it in your ISP as a server farm. Maybe you're a university that's basing your custom distribution of Linux on Fedora. Maybe you're an animation studio that's using Fedora to make movies that make hundreds and hundreds of millions of dollars around the world. But the project itself has to have a small set of goals that define both what it is and what it isn't so that as decisions get made and as resources get prioritized, you can make progress toward those specific things. And Fedora, I think, is at a point in its life where what is important to it is increasing the size of the open source development community around the world. We need distros, and we need companies that want to spend a whole lot of money marketing Linux and getting as many new users into Linux as possible. And I applaud folks like Ubuntu that seem to have done a great job with that. Fedora does not compete directly in that space. I want you guys to picture a big, giant, reverse pyramid that represents the potential population of the open source community. And at the very top of that pyramid, you've got a whole bunch of basic users who are simply consumers of a Linux distribution. It doesn't matter which one it is. And then of all of those users, some percentage, and I would argue to you that it's a relatively high percentage, if they run into a problem, they're going to go on to Google or whatever, and they're going to search for their error message, and they're probably going to find some answers, and they're going to solve their problem. Great. Until that moment in time, everything that user has done has made their computing experience easier only for them. There's a pretty big gap in this pyramid to the next step down, which is the moment in time in which a user takes an action that turns them into a participant in a community. And that gap is where Fedora is really trying to focus, in my opinion, right now. There's other Fedora contributors here. Maybe they have different ideas. But I think that's kind of where we're focusing. So getting those users who are solving problems for themselves to do something as simple as file a bug report or post a little how-to of how they've solved their problem on a forum, the moment that action gets taken, all of a sudden a user has become a participant. And that is the big gap that you have to achieve. And from there, it becomes easier for people to go from being someone who files bugs to someone who submits patches, and someone who submits patches to someone who wants to maintain packages. And then you've got people who are building infrastructure, writing new applications, developing new features. Once you make that first leap, it becomes so much easier to pull someone completely into the open source ecosystem. Fedora strives to be the place where that kind of work can happen and where people who want to fully engage in the apprenticeship that is open source development can get into a community that allows them to succeed. We'll talk about that a little bit more in this presentation. So I will very quickly say, when I introduce myself, you'll notice that nowhere in my introduction did I say that I am a incredibly smart, technical person. So if you've got questions at any point in this speech, shout them out. If it's technical question, I'm probably going to turn to Tom Calloway here, who's a Fedora engineering manager or one of the crew in blue and let them answer your question about exactly which internal driver it does, blah, blah, and blah, blah. But feel free to shout out any questions or comments at any point in time. For the rest of this talk, I want to break it down into four categories. And these four categories are ones that were chosen carefully by our Fedora marketing team. And we call them the four foundations of Fedora. Again, I didn't realize the room was as big as Wembley Stadium, but I've got some stickers here. And if you want a Fedora sticker, you can come to the Fedora booth and we'll give them to you. They say on them freedom, friends, features, and first. Yeah. Those are the four foundations that we believe represent kind of the core philosophies that Fedora is built upon. So I want to talk a little bit about each of those four things, kind of say what it means to me, and then give some examples of what's going on in Fedora right now or recent achievements in Fedora that relate to those four buckets. And hopefully, you'll find it interesting and we'll go from there. The first one that I want to talk about is the Freedom Foundation. Because I think that's the most important, right? We're all here because we believe very passionately in free software. And Fedora's commitment to free software, I would argue to you, is among the strongest of all of the Linux distributions that are out there. And Red Hat's commitment to free software, oh, hold on. Red Hat's commitment to free software is that Red Hat recognizes that it is building its business around a giant community. And if Red Hat wants to be able to look people in the eye and say, we are the leader of free software in the world, then Red Hat has to make the commitment to ensure that its participation in the free software community is genuine and honest and strong. So we'll briefly talk about licenses and spot you can jump in any time if I say something wrong. Fedora is very active in making sure that all of the code that goes into its package, right? All of the packages that we put into the distro and all of the upstreams that are building new code are contributing code to the distro. We want to make sure that all the packages in Fedora have a free software foundation or OSI compliant license and spot himself spends a large amount of time working with upstreams that want to get code into Fedora or other distros where there are licensing problems and tracking down the copyright holders, helping explain to them where there might be problems in their licenses and getting licenses changed so that the free software commons that everybody can draw from is as large as possible. That's all in there. I'm not really going to go into licensing too deeply here, but I wanted to make it clear to anybody who doesn't know that Fedora holds a very strong line about free software and we're very passionate about it. The more interesting aspect of this freedom foundation with regards to the Fedora distro as I see it is the manner in which our community participates in building the distribution. If I had stood up here three years ago and tried to give this talk, it would have been littered with a to-do list of all of the reasons and all of the places where the Fedora distribution isn't as open and isn't as community friendly as it needs to be. But today, we can stand up in front of you and we can proudly state that about two-thirds of all the packages in the Fedora distribution are maintained by people in our volunteer community. Anybody in the world can be a maintainer or a co-maintainer of any package in Fedora. All they have to do is be able to step up and do the work required to maintain and own that package and they can do it. What's also interesting is that the release engineering processes and the infrastructure processes around Fedora have been specifically architected over the last couple of years and re-architected in order to make the community engagement as vibrant as possible. Fedora is the perfect example of a distribution that eats all of its own dog food. The build system that we use to take our code and create source RPMs and take the source RPMs and create binary RPMs is completely open sourced. I guess Joe was talking about the open Susie build system a little bit. I'll talk about the Fedora one. It's called Koji. It was written in conjunction with some Red Hat employees and some community members. We took the good parts of the Red Hat build system, the stuff that was generic, the stuff that wasn't needed for Red Hat Enterprise Linux. We streamlined it. We worked on it with our community. We added features. We built a new build system that anybody in the Fedora community can use. It was so good, in fact, that the circle goes round and round. And Red Hat is able to take the Fedora build system that it helped invest in building and, over time, replace more and more bits of the old crappy Red Hat Enterprise Linux build system so the ecosystem shows. Fedora, I've got to go put my Red Hat on again. By investing in the Fedora community and doing everything that we can to put the ability to innovate into the hands of the most passionate and most educated people in the world who care the most about it, Red Hat gets more efficiency in its own processes. Being a good citizen and a good contributor to the open source community repays Red Hat tenfold for its contributions. Moving on from the build system, because Jiren van Moywen is in the room and I like to embarrass him, is the system by which we actually compose the distribution. You've got to take a big pile of RPMs and create some sort of ISO out of it that can do something. And again, we see the openness of the community really coming into play here because the main compose tool that Fedora uses is something called Pungy. It has a nice API. Pungy is great, by the way. Simple command line. You really almost don't have to know anything in order to successfully run it. You just say, build me a distro and you give it a kickstart file or a pile of packages and it'll go ahead and do it. And that's what the Fedora release engineering team uses to build all of the different versions of Fedora that we released to the world. But it's got a really nice API on it and it allows community guys like Jiren and a group of people called the Fedora Unity Project who want to build upon that work to make a graphical interface on top of it. So someone who wants to just go through a wizard and build their own distribution can do it. A guy named Bill Nottingham, one of the most senior engineers at Red Hat and in the Fedora project, did a lot of work in taking all of the logos and the trademark issues around Fedora and getting all of the trademark and logo stuff into one package that can easily be removed and replaced with a generic package so that people who want to create downstream derivatives of Fedora can do so as trivially as possible. Because all of the tools to build the distribution are open and because we've gone the extra additional step of making it very easy for people to rebrand and reuse the distribution and remix it without having to worry about legal and trademark issues, we see a very vibrant release engineering and custom engineering community sort of building up within Fedora. On the enterprise side, I'm shadow man, I'd like to sell you $10 billion worth of software. On the enterprise side, we see a group of people inside of Red Hat using these ideas in a project called Fincrust that's sort of a Linux appliance building tool. I've now reached the level of my technical knowledge about it. But again, ideas that are happening in the community, ideas that came from the community that we supported bringing value back to the business which allows us to continue to support the community. Well, that's what I wanted to say about the freedom point. One fourth done. Any questions or comments? Anything anyone want to add from the Fedora world? The next of the four foundations that I want to talk about is we talked about freedom, now I want to talk about first. And to me, first represents innovation, okay? And it is the goal of Red Hat and the goal of Fedora to be the most innovative community in the open source development world. And we believe that we are. So this is where, this is the part of the speech where I might need some help from my friends. I want to talk very briefly at a relatively high level about some of the technical advancements that are seen all over the Fedora ecosystem, or I'm sorry, all over the open source ecosystem that have come from the Red Hat and Fedora community, right? The first one I want to talk about is a new ex-server project called Wayland, right? This is relatively recent in the last, I don't know, maybe nine months or something like that. You see a lot of it in Fedora 10, if you've used Fedora 10. We've got a program in there and it's when the operating system starts up, right? Right after you see the BIOS, there used to be a program called Red Hat Graphical Boot, it was the little Fedora logo and it would switch from all the different modes as all your services start up, it would run down the whole thing. That's the old world. We replaced that in Fedora 10 with something called Plymouth, right? Which takes advantage of the kernel mode setting feature that some of the newer chipsets have, and Wayland, which is the ex-server that takes advantage of this, so that the entire boot process now is significantly faster, more streamlined, right? There's a feature in Fedora 10, Fedora 10 and Fedora 11. It's a continuing project through the current two releases of speeding up the boot cycle. The goal, I think, for Fedora 11 is to get a 20-second startup, right? From turning on the power button to GDM, you're ready to log in, everything you need to go in 20 seconds. All of the development behind the scenes on Wayland is going into that and we're pretty confident that because we're doing that development openly, it's going to be something that benefits the entire free software community, right? Another thing that Red Hat and Fedora have invested significantly in is development around the Nouveau driver, right? The reverse engineering of the NVIDIA 3D drivers. And again, these are developments happening in the core functionality of Linux, right? X, or I guess, would it be fair to say that it replaces X? I don't know. No, around X. That are going to benefit the entire community, right? Red Hat and Fedora are leading the way here. We think that's a valuable thing to do. The other place that I think it's a useful place to talk about, because we see it being used more and more in all of the infrastructures that different Linux distributions are rolling out, is virtualization, right? It's been a major thing for a long time, ever since VMware got us going years and years ago. But the inevitability of open source, right, always said that VMware's proprietary model would eventually be replaced by an open source community, which can do these same things better and faster and for less money. And here we are in 2009, where we've got a progression. If you look at the history of Fedora from Fedora Core 4, all the way here to the alpha of Fedora 11, you've got a long progression of progress in the virtualization space, starting with the work that was done on Zen way back in the day, right? But let's all be honest. What's the leader in virtualization right now? It's KVM. So what does the Fedora model make possible in something deeply technical like virtualization where you've got different competing solutions even within the open source ecosystem? How does Fedora help if you're a developer? How does Fedora help you make sense of what's going on here? And work with technology that changes very quickly? Well, Red Hat has invested a lot of its own engineering resources and built community around various virtualization projects like Libvert, right? Which tries to consolidate a lot of the virtualization calls that need to be made into a basic API. Vert Manager, which begins to give you a graphical interface to use and manage different virtual machines, right? And moving on even further, in the recent history we're starting to see in the Fedora project called Fedora Cumulus, which begins to take advantage of the virtualization technology that exists and build sort of an infrastructure cloud which we're going to use both in building the Fedora distro so we can make it available to a huge community of hundreds and hundreds of people who are building thousands and thousands of packages. We can manage all of this stuff and because it's been built with an abstraction layer on top of it, right? As certain people go and use KVM, we're capable of supporting that. As certain people go and use Zen, we're capable of supporting that and I've got to put my hat on. The value for Shadowman is that you've got a huge community of developers who are playing around with these competing virtualization technologies who are helping to ensure that the manner in which Red Hat is building APIs and general means of using virtualization is sane and makes sense and can support something as robust as the Fedora infrastructure so that by the time we take all of this virtualization technology and stick it into Red Hat Enterprise Linux, at which point in time we're committed to support it for seven years, we're so much further down the road of knowing that it's stable, that people want to use it, that the engineers who work for our customers aren't gonna look at their managers and say, we bought this crap, right? It's gonna be something that our customers are able to use and that we're capable of supporting. In the Fedora space, however, what's exciting if you're a developer is that you've got a six month release cycle, right? In the United States last year, you might've heard we had a presidential election. Over a billion dollars was spent by two guys, campaigning over the course of, I don't know, it took them three years, right? So that one person could get elected president. In Fedora, we have a technical election every six months, right? Every time we throw out a new distribution of Fedora, everything is kind of up for grabs, right? And it's a chance to look around the open source ecosystem and decide what are the most compelling pieces of software that solve various different problems and pull them into the distro, right? And the cost of doing this, because we allow the community to lead us and the developers to show us what is most exciting, what works well, where is the mind share going, right? Because we allow the community to sort of be that massive electoral body, the cost of developing an enterprise Linux distribution, the cost of developing an enterprise Linux distribution that anybody can use drops very far, right? The price of failure is lower. And because of that, the entire open source community, because let's not forget, there are derivative downstream distributions of Red Hat Enterprise Linux as well. The entire open source community gets access to enterprise quality software in theory at a price much lower than what it should cost for proprietary software companies to do the same thing, right? And that larger circle that isn't just Red Hat, but it's the entire open source community working together because everything is GPL and everything goes back upstream, that gives us the energy that we need, right? To continue to threaten and push the proprietary companies out of business, hopefully. I could ramble on a little longer here, but instead I'm just gonna name a couple of other projects, right? Upstreams that have been adopted across various Linux distributions that have been led by Fedora and Red Hat to try to underline the point that, again, our ecosystem is trying to attract the developers who are interested in getting their hands dirty and doing all sorts of new things, right? So you've got Pulse Audio, you've got Network Manager, you've got Package Kit, you've got Debus, you've got HAL, which is now being replaced by Device Kit. These are all things that are way over my head, right? But these are fundamental pieces of the Linux system where innovative work is being done within the Fedora community and where we're looking for more and more developers to help us test things, to help add features, to help really be a part of it and cut their open source teeth in our ecosystem so that we can all then go forth into whatever companies we go to and be more equipped to spread the open source message. Any questions, any questions that I say anything wrong in the technical world that anyone wants to correct me? Yeah, is that, Joe? Well, yeah. Is that, is it putting the companies out of business as a goal or just having the alternative? No, I think, yeah, I was a little careless with my words, I suppose. I think the goal is to show that the open source model is clearly superior, right? And that you will see as new companies and new technical companies begin, they're gonna say, well, it's blindingly obvious that adopting an open source model and investing in and joining the community that already exists is gonna make us so much more efficient that sooner or later, the guys that are graduating from Harvard Business School are gonna assume, yes, you know, I have to build my community in an open source way if I want to be relevant. I think that's a better way to phrase it. And again, it's a good question. And I don't wanna drag myself too far off on a tangent, but a point that I used to make in an older talk that I gave is that for a lot of us developers, right, open source participation is a very philosophical choice. We believe very deeply in free software, and that is why we do it. But when we talk to the guys that wear their suits and ties, right, they understand that and they respect that and they appreciate that, but we also need to be able to show them that it's a sound business investment, right? That they're going to make more money and their company is going to be more successful if they embrace open source. And that's part of what the Red Hat Community Team's job is, right? Within the Fedora space, we've got people who drink the Kool-Aid and who believe it very passionately and they're gonna participate because it's in their blood, right? And within the Red Hat space, we've got a lot of that too, but we wanna make sure that we are consistently reminding everybody within the company from the top all the way down to the bottom that what makes us a special company and what makes any company that's an open source company special is because not only do you acknowledge what open source communities can do for you, but that you buy in completely, right? And that you don't go halfway, right? You don't just throw your code up on a website and say, well, I guess we're an open source company, right? If you're not going to build community around it, if you're not gonna actively manage those communities and if you're not gonna try to allow those communities to leverage the investments and resources that a company wants to make, then you might as well not do open source at all, right? I think it would be better to go, it's an all in or all out proposition, right? Either you gotta be an open source company and go all the way or you should just not bother at all. If you go halfway, I think you give up all benefits of, you give up all the benefits that in theory having proprietary IP could give you and you get none of the benefits that open source could give you. So while I'm on this topic of community, I'll move to the third of the four foundations that represents Fedora. We talked about, oh, my mind is jet lagged. We talked about freedom and we talked about first, now I wanna talk about friends because so many of my friends are here in the audience, right? This is the less technical part of the talk. I already talked about the fact that about two thirds of the packages in Fedora are maintained by our community. I wanna talk briefly about a couple of other pieces of the Fedora community that I think provide the ecosystem around the distro that folks might not know about. Joe did a little bit of this in his talk as well because the distro is great and it's critically important but the Fedora project is so much greater than just the CD that you can go get at our booth. The Fedora infrastructure team, I think is a shining star in our community, right? Because it embraces perfectly the community building tactics that we have tried to perfect and implement over the course of the last few years, right? It is almost entirely volunteer driven and it is also entirely open, right? Every piece of software, every single piece of software that's used to keep the Fedora project running is completely open source. We won't use anything that's not open source. We completely eat our own dog food. We've got a system administration team that basically follows the sun, right? We've got volunteers all over the world who are part of the Fedora infrastructure team. If a server goes down in the middle of the night in the United States, in theory, there ought to be someone somewhere where it's daytime for them who's gonna be able to respond to that message quickly and in theory, no one should have to wake up in the middle of the night in America. In fact, our infrastructure itself is co-located around the world and we've recently in the last year so built up a pretty nice data center in Germany that helps us make us kind of fault tolerant, I suppose. And the Fedora infrastructure project, from the perspective of someone who wants to participate, I think is an excellent place to get started because we've seen some of our recent success stories of really brilliant new contributors to the project come through the Fedora infrastructure team, right? There's a guy named Ricky Zhu who's now 18 years old who started contributing to Fedora when he was 16. He came in through the Fedora websites and the Fedora infrastructure project. He's just started his first year of university and he's one of the most important people in the Fedora infrastructure project. And I know there's stories like that all over, the open source community, but I think it really represents for any of the people in the audience who are still in school and looking to get their first job in open source or looking to make the first step of what's gonna be a long career in open source software, I think it shows how these communities, if they're built right and if the leaders understand that they have to think about building community and not just doing things themselves, it shows the incredible potential for someone at a young age who otherwise wouldn't be able to get an opportunity in some normal sort of corporate situation to build up such a huge amount of experience and an authority that when they're ready to get their first full-time job, they're gonna be years ahead of most of the other people that they're competing for. And I think that doing that encourages such a positive feedback loop that it's important that all communities work on that because it will continue to increase the size of the open-source workforce, right? The last thing I wanna mention on this sort of education or young contributor point of view is we have a project in Fedora called the Fedora Scholarship. And we give it out, this is the second year that we're giving it out and the application deadline for it in fact is sometime next week. I think it's the 15th of February, next Friday I think is what it is. If you are somebody or if you know somebody who is very active in open-source, preferably very active in Fedora but very active in open-source, who is starting their first year of university later in 2009, so like the fall of 2009, encourage them to go onto Google and type Fedora Scholarship and read about it because this is something that we're gonna do every year and we want to basically give an incentive for people who are right at that point in time where they're gonna start their university career who have shown interest in open-source. We wanna help them pay for their university bills and encourage them to stay in the open-source community as they go through university. So I wanted to kinda make that little sales pitch. The last thing I wanna talk about in the friends section of my talk here is an area where I think Fedora needs to use community to improve over the course of the Fedora 11 and Fedora 12 release cycles and it's one of the goals of Fedora all the way from our Fedora project leader Paul Frildes on down and that is to improve the quality assurance portion of Fedora, right? One of the things that we have struggled with in Fedora over the last couple of years is this challenge of getting stuff done versus building capacity that allows more people to do stuff, okay? Building the capacity is fundamental to community, right? You've gotta build a infrastructure in which people can participate before the people can participate. If you just have one person trying to do the job all by themselves, they're quickly gonna get overwhelmed, they're quickly gonna burn out and they might be the most brilliant person in the world and our QA people are. But they're going to reach a limit very quickly where they simply cannot work 24 seven and get everything done that needs to be done. So in recent history, I would argue to you that Fedora has fallen into this trap over here, right? Too much time needed to get the basic QA for the distro done, not enough time available to build a community around Fedora QA. It's improved, right? Over the last year, so we've had some people that have stepped up and rebuilt Fedora's bug triage community. We've had a guy named James Laska lead a series of Fedora test days, right? That have been very good. We've had basically an IRC channel and a specific topic for each of these days. We've tested things like PyKickstart, the EXT4 file system in Fedora 10. We tested all the full disk encryption and its integration into the Fedora installer, right? So we've seen enough of the seed of the fact that with the right ideas and a little bit of time spent, you can build a community around QA. And this past week is very wonderful for this project. Guy named Adam Williamson, who used to be with Mandrieva, joined Red Hat and his job is going to be to build a community around QA, specifically in the Fedora space. So that's another area critically important that there's a lot of opportunity for participation. So I wanted to be sure to mention that. The last thing that I wanna talk about is the last of the four foundations is features. So we've got freedom, we've got first, representing our innovation, we've got friends, representing our community, and we've got features, representing what on earth have we actually produced for you? And yesterday, I think it was yesterday, now Thursday, two days ago, Fedora 11 Alpha was released and I wanna quickly run down some of the cool things that are in it and then encourage you to go and download it. I mentioned this earlier, the goal in Fedora 11 of getting startup time down to about 20 seconds, right? The first cut of that, that we started in Fedora 10, continues in Fedora 11. A lot of the way they're going about doing this is trying to identify the places where IO really tends to block during the boot process and eliminate that if you wanna know more AskSpot. Again, the first pass of device kit, which as I understand it is sort of a rewrite of how the hardware extraction layer is in Fedora 11. EXT-4 is replacing EXT-3 in Fedora 11, little crufty, EXT-4 is some better performance. There's also the first parts of butter FS, I guess that's how I say it, I don't know how you people say it, but if you wanna try that, that's starting to get ready in Fedora 11. The feature that I like the most is something that we call Presto, what it really means is Delta RPMs. So let's say OpenOffice, for example, releases an update and you gotta download, you run YUM update and you've gotta download some 20 megabyte RPM, but the amount of code that actually changed in that 20 megabyte RPM is only a few kilobytes because it was just one security update. The Delta RPMs feature is gonna make that possible so that you don't have to download the entire new update and have 90% of what you download be exactly the same bits that you already have. So if your bandwidth isn't great, I think that's a really compelling feature. It also, it just feels more efficient and I kinda like that. So there's also a lot of you guys, I know I discovered this a few years ago from the Windows side, there's a project called MinGW, which allows you to do cross compiling of applications so you can write your Linux application or you can write your application in C, you can use GCC to compile it for Linux and instead of having to rebuild it on a full Microsoft stack or something like that, you can use MinGW to build the Windows executable of a free software application that you've written completely in your Linux system. So if you're an engineer who has to do, who has to build applications that run on multiple platforms, here's a way to do it without having to spin up an entire Microsoft stack. Python 2.6 is in Fedora 11, as well as GNOME 226, KDE 4.2 and XFCE 4.6, the KDE and XFCE are a release candidate or betas right now but we're also still a few months out from Fedora 11 finals so all of those projects will continue to mature over time. In conclusion, well, before I conclude, are there any questions? I can't hear, that's why I'm walking toward you, not to throttle you, yeah. Thanks, one of the questions I have is what you guys are doing specifically around KDE 3.5 and KDE 4.0 series because with OpenSUSA, we've had a large number of KDE users who want to stick with three, we've had a large number who want to be on the cutting edge so I'm curious what Fedora is doing to solve that or what are you guys doing for them. Spot, do you want that? Can you answer that? You probably could answer that better than me. We have no intentions to going backwards and providing the three environment from a compatibility perspective because we feel strongly that at this point, KDE 4.2 is reasonably featured comparable with the 3.5 series and where there are gaps, we'd much rather encourage contribution to fix those gaps than have people just stay stagnant on the old stuff. Right, right, right. Cool, well thank you. Michael Thiemann, who basically created Cygnus, is here somewhere, so yeah. So I'm gonna wrap it up, I'm almost done. Oh, yes, you go, then I'll wrap up. You talked about Zen and KVM earlier on. Some of us who are dealing with virtualized environments preferably would like to have a common environment. For instance, Zen in Solaris goes with Zen to a certain degree in parallel to Solaris containers. So the question is, is Red Hat and Fedora going to support Zen in the future or are you going with KVM on it? Well, let's be very clear. Red Hat already supports Zen in Red Hat Enterprise Linux and is therefore committed to supporting it until the end of life of Red Hat Enterprise Linux which is seven years or whatever. Fedora never makes a commitment to support anything for longer than 13 months, right. This is part of Fedora's value proposition, right. We believe we can be innovative because we don't have a large legacy set of stuff that we have to support. So Shadowman says, I would encourage you to give me as much money as you possibly can to support Zen because I will do it for you. I will sell you all the Zen support you like. Fedora says, we're gonna take advantage of the fact that we've got VertManager and LibVert and APIs that allow us to switch out any virtualization from Zen to KVM or anything else that we want underneath and the user never has to know about it and Fedora is gonna continue to do the best of what exists today. So that's all I got for you. All right, I am out of time now. So let me quickly say a few words about our Fedora community in Europe because it rocks. If Jono Bacon were here, he would say it's metal. We got a great group of guys here and girls that keep Fedora running, that operate all of the major events and the next big place that Fedora is going to be is gonna be at Linux Tog in Berlin in June. We're going to hold our annual European flood con there. It will be in conjunction with Linux Tog. It's basically going to be the biggest Fedora event of the year in Europe and I would encourage all of you to come to Linux Tog, there you go. If you want to join our community, we would love to have you. Feel free to come by our booth. Feel free to, if you're ready to become a contributor, go to join.fedoraproject.org. It's a trivial process to get started, get your little wiki page up and have at it and I thank you very much for your time and attention. Thank you.