 Good morning, ladies and gentlemen. I would like to welcome you to this, the final day of DEVCOMF for 2016. Our opening talks this morning is the Fedora keynote and without any further ado, I'd like to introduce you to Denise Dumas and Matthew Miller. Thank you. I'm impressed to see such a fabulous turnout at nine o'clock on a Sunday morning after the big party. Wow. It's nice to see that we have such a dedicated Fedora crowd. It's a beautiful thing. So probably many of you know me already. Some of you work for me. But I'm Denise Dumas and I am the VP of engineering at Red Hat who is responsible for Red Hat Enterprise Linux. And I am going to try to explain, at least from the Red Hat perspective, what the Fedora and Red Hat relationship can look like. So I did this talk first last year at Flock and we called it What Does Red Hat Want? And we had a lot of fun with it. But since then, things have morphed a little bit. So I wanted to do the updated version of this talk. And once again, I'm going to start by talking about Red Hat's mission statement. So the first thing that you need to understand about Red Hat is that things like this, they don't come landing down on everybody from up top in the management layer where the air gets thin. This is the stuff that we work on together. And Red Hat's mission statement to be the catalyst in communities of customers, contributors, and partners creating better technology the open source way. If you work for Red Hat, you know that every word of this statement was debated and discussed and agreed upon before it all came together. And we were very careful to make sure that all the nuance was correct. You notice catalyst. So when you think of catalyst, you think of in the middle of the mix helping it bubble and making the reaction go faster. Not following, not leading, in the middle and helping everything move along ideally together so that in the end we can gel lots and lots of ideas into a fabulous technology that everybody wants to use. So Matthew and I were talking about the Fedora mission statement, which as it turns out is remarkably complimentary. Matthew? To lead the advancement of free and open source software and content as a collaborative community. Yes. I think we're in sync. So there's another part that goes along with the mission statement. And this is one of the things that they teach you in manager boot camp. So you become a manager at Red Hat. And right after the lobotomy, they take you out and you go off into the smoke filled hut. And you get, you know, and then we sit around and we say, all together, no margin, no mission. No margin, no mission. The reason why we can be the catalyst and support the growth of open source companies is because we're able to stay in business. And the way that we stay in business is by listening to our customers and bringing that back to communities and helping them understand a different side of the world. And what the enterprise values, what Red Hat knows, is stability, scalability, performance, security, certification on widely available hardware, manageability, and increasingly, the ability to develop and deploy your applications faster and faster and faster. And that's huge. And that's what we keep hearing. So as you see from the mission statements, Red Hat also shares community values. And I can do this by heart. Freedom, friends, features, first. And I'm adding another one. Faster. I mean, I hope you guys know, well, I'm sure, you know, everybody from Red Hat, right? If you work for Red Hat, chances are you're running Fedora, you're working in Fedora, you're packaging, you're submitting patches, right? We are part of Fedora in a big way. And we want more people to be part of Fedora. Fedora can mean innovation. So here's something I want to point out. If you saw this talk at Flock, at Flock, I said this as Fedora means innovation. There's a subtle difference here. Fedora can mean innovation. Fedora needs to be innovation. Faster, faster, faster. Right? If there's a theme. Fedora gives all of us a chance to participate in distrawide experiments. And it's a place to experiment not just with merging together all the upstreams, but with experimenting about the distro itself. What does it mean to be an operating system in 2016? What is it going to mean to be an operating system in 2020? Because that's what we're looking at now. Where are we going? Faster, faster. So two releases a year, that used to be fast, right? That's not fast anymore. We want to be moving continuously. We want to be DevOps. We want to be able to have the kind of infrastructure that lets us do continuous development, continuous integration, continuous validation, continuous delivery. Faster, faster. And we need to be a place where we can try out different approaches because, you know, you never get it right the first time, right? It's always going to be, well, let's see what they think about the prototype. And that's always been a value of Fedora. Let's give it a shot and see what happens. That supports Fedora. And honest to God, I know Red Hat supports Fedora because a lot of this comes out of my budget. So you know that competition for resources thing, right? But think, 35 full-time people, that's more than a lot of startups have. If you want to put it into perspective, we have 35 full-time Red Hat people working in Fedora across release engineering, across the IT infrastructure, and across the QE organization. And lots and lots more Red Hat people participating in Fedora as packages, as maintainers. So everywhere you go, you scratch a Fedora person, we hire from the community, right? Sometimes that's a problem. We hire half the community. But we support Fedora everywhere we can. The hardware budget, like this year, I know, 150K into new hardware for Fedora, right? I know that. That was our budget. Thank you, Paul. So why does Red Hat support Fedora? You've all heard the expression, there ain't no such thing as a free lunch. We support Fedora because our missions align. Fedora is a way that we can advance the cause of open source software development. It's a wonderful place for that. And Fedora is also a place where we can jointly experiment with ideas for our next product. And guess what? That's an operating system, right? So how do we become more innovative in the way that we do operating systems? And it's true, there ain't no such thing as a free lunch, first law of economics. But the other thing is that there is no big secret agenda. It's not, oh, I couldn't possibly share that with Fedora. We're Red Hat. We're open. We're transparent. It's a core value of our company. Hey, here I am, right? The VP of Rail Engineering standing up here telling you, yeah, here's what we want. And you know what? I'm not very subtle. You know? It's been a fight. I want you guys to meet Josh, who is the Atomic Community Coordinator. Josh, take a bow. Atomic helps us get there. It gives us a way to experiment with where the operating system is evolving to. The next generation of the distro, here's what we think. Here's what we're seeing. Automated, modular, smaller, faster. Even more security-hardened? Easy to update from release to release. These are all the things that we think are going to be important, but these are also big, vague, blobby words, right? What does this mean in reality? Fedora is where we want to learn. With the Fedora community, because we think that this benefits the Fedora community, this keeps Fedora interesting and relevant and fun to work in. We think that it gives Fedora life, right? This is not your father's distro. So what Red Hat promotes for Fedora, what we want to see is willingness to change and to grow. You guys all know about evolution. You evolve into the perfect match for your evolutionary niche. And then what? Disruption happens and you become a statistic. An unfortunate statistic, right? The Brontosaurus. Well, we don't want to be the Brontosaurus. We want to be moving and growing and staying out ahead. We want Fedora to be a community that considers the status quo to be a hurdle, something to get beyond. We want to be constantly adapting, constantly evolving. Fedora is never going to be done. If it's done, it's dead, right? So let's keep it moving. Make it fun. We want more people to be able to participate in Fedora. We want to make Fedora attractive to developers, to users. We want the Fedora desktop to be something that people adore. We're all trying to support that. Matias, you know, we want that desktop to be fabulous. What Red Hat plans is to help Fedora move faster. There's a theme here. And the way that we think we can help that is to communicate clearly, clearly what we want with one voice as best we can. We want to have a positive, constructive dialogue. We don't want to be fighting it out on Fedora Developers, right? You don't want your children near Fedora Developers some days, right? We want to be open and clear about our desired technical outcomes. We understand when things, that things are not always going to go the way Red Hat wants, right? Because Red Hat doesn't have a monopoly on good ideas. Community. We're with it. And the other thing that I hope you know about Red Hat is that we're willing to put our money where our mouth is, and we send patches. But here's the thing. If you don't evolve, you get disrupted. Disruption is coming. Red Hat smells the smoke. And we plan to be out in front of the fire. And we want Fedora to be with us. In fact, what we really want is for Fedora to be outrunning us. But Red Hat is not going to be, we're going to be part of the disruption. We're not going to be a victim of the disruption. So we're going to be moving faster, faster, faster. And my challenge to Fedora is to make us look slow. So I'm going to turn it over to Matthew Miller with a thought about what Red Hat wants to create the operating system for the next 10 years with the Fedora community. Thank you. Okay. Mic'd up. Okay. So segue from Denise's talk. First, my talk comes in two parts. First, I'm going to talk about the status for the last year and so and where we are and where we've been. And then I talk about what's coming over the next year. So keep what Denise is talking about in mind and then let's kind of look backwards for a little bit before we look forward again. We really had a great year this year. And we don't always get a lot of press because Linux distributions are not very exciting to the mass audience anymore, which is part of the theme of why we need to do exciting things that make people notice us. But the things we did do, we did really well and the press we did get was glowing. This one in particular I'm proud of and I think we all should be proud of. The register, if you don't know, is basically a tech tabloid and they kind of live off of snark. Usually when they say something about us, they've got a lot of, you know, negativity lurking behind the surface. But this isn't when they did their Linux distributions a year in review basically glowing and not just about the product we made, the thing we produced. But actually it's about us as a community and where we're going and our strategic vision actually meaning something. So I think that's really awesome. I'm proud of all of us for this. So we got fedora by the numbers here. First I've got numbers about downloads and things like that, connections. And this I've always promised to start with a scary dinosaur, not because the numbers are ancient but because big sharp teeth. There are a lot of caveats about how these numbers are gathered. We don't do any invasive metrics. We don't measure all your keystrokes and send them to our server and figure out what's going on. We kind of observe, yeah, right. We observe what's going on in the wild and we try and extrapolate from that, which means that when we've got charts here with the y-axis, with numbers on them, we can only guess what those mean. When it says 100,000, we don't know if that means that that's 100,000. It's all there are or if it means there's only a tenth of people are being counted. So take that with a grain of salt and also the way the metrics are being gathered basically is very dependent on the network topology. So we basically count IP addresses. So if someone is at a university where everything comes out through one proxy, we count everybody at the university one time per day. And if somebody is moving around between a lot of IP addresses, we count them multiple times. We probably think that it's undercounting. And also, the way we count our connection updates is basically really loaded towards parts of the world where the internet is always there and cheap and something you don't think about. So we are definitely undercounting users in countries where that's not the case. So here is the first really hard to read slide where we have a chart, basically a peak for this is the connections to our update server for all of the Fedora releases. And in the interest of time, I'll just skip that one to the simplified view here where I have lumped them together. And so basically, I kind of arbitrarily chose different Fedora releases to put into a lump. We start over in 2007 here on the left side. And that's like Fedora Core 6 and 7. 8 is the light blue thing all by itself. That was a weirdly popular release. We think it's probably Amazon EC2 because it was a very popular early available image. And then when we got to 9, it didn't work in Zen anymore. So people stayed on 8 for a very long time. You can see it continues to this day. Then we had a nice period of the purple age of Fedora where we had growth. And then we had Fedora 15 came out and disconcerting things happened like lines going the wrong direction. There's a lot of speculation and finger pointing about why that might have happened. I don't want to dwell on that too much. There was a lot of change. We had the first system D. We had the first GNOME 3. And well, I think those are both amazing technologies. The first releases were shocking, I would have to say, and a bit rough. So we've done a lot of polishing over the years. But it's a hard message to get people to come back and look. It's awesome now. So we are on this downward slope. Then we started the Fedora Next initiative. When we got to the 10-year mark, basically, we said, OK, where are we going to go for the next 10 years? How are we going to do this? And we've been seeing that number going down. How can we turn that around and we could go up again? So the yellow one is Fedora 20, which was a year-long release when we basically spent the time to retool and figure out where we wanted to go next. And then the green is where we started with our strategy with the three separate editions, Fedora Workstation, Fedora Server, and Fedora Cloud. And I think the line there kind of speaks for itself. I think that has really worked. So yes. And this is an even more simplified version of it, just showing you the total there, basically, and then the peak for each release as we go. And I had to put a caveat on Fedora 23 here because the first time I showed it to Denise, she was like, what have you done? What happened to Fedora 23? The thing is the peak for each release comes right before we release the next release. It's often the day right before we drop a new release. The previous release is the most popular we'll ever get, which I think is interesting with our six-month cycle. On the one hand, six months is not fast anymore. On the other hand, it doesn't seem to be really that that six months is what's driving the user adoption. People aren't waiting for us to have a new release, and the new releases aren't making people be like, oh, wow, I was waiting for that, and I'm going to, now I'm going to be on Fedora. So I think that's a thing for our marketing to look at if we can make our marketing more driven about what Fedora is and less about what we've got in each new release, which is kind of an easy way to generate some marketing things. But again, like I was saying with the press, like the press isn't biting very much anymore when we say, oh, look, we've got Ruby 2.13, and we've got this kind of thing. Like, that doesn't make a good news story. And so, anyways, the, the, we've updated stuff cycle isn't really, isn't really driving, a driving factor. Okay, more numbers. This is downloads of the actual installer from GetFedora.org every week. And you see here the access is basically, you know, goes up to 120,000 a week or so on the release days and release weeks. And overall, the size of one blob here is about 1.2 million downloads per release, which is a pretty nice big number. And another thing where, you know, we're looking at the Y-axis. What does that mean? You know, do you download it and then you do it install fast and install 100 systems. You download one cloud image and install a million systems, you know, or a million systems every day over and over again. Or do you download it and throw it away? We don't know. But that's downloads. So, of those downloads, this is the split between the different editions. So, Cloud Server and Workstation here. And you can see that the blue thing here is Fedora Workstation, which is, you know, desktop, the, the, the target goal is basically to make this really appealing to software developers. That doesn't mean that we're not appealing to everybody else, but we want to make sure that software developers find this to be the place to go so that when we go to a Linux conference, we don't find everybody running a Mac. We want to have, they should be running Fedora instead. If they've got a Mac, they should have Fedora on it. And we also want to branch out beyond those Linux conferences to the wider world of developers and try and get everybody who's working on the new next cool startup thing to be using Fedora to make it. That's the goal of Fedora Workstation. And anyways, it's our most popular edition by far. But we do really have a really strong server, Fedora server download, which I think is important, because, frankly, Fedora server is more aligned with what Red Hat needs for the Red Hat's major server audience and, you know, the RHEL product. Obviously, there is a RHEL-based desktop as well. But, you know, server is the bread and butter for Red Hat. And so we want, as part of our collaboration, to be a good upstream, we really needed to have something that is more directly aligned. And it's also fits the traditional Fedora contributor audience, where we have a lot of sysadmins who came into Fedora, you know, 10 years ago and still tend to have a lot of, when I go to Lisa, the large installation sysadministration conference, so many sysadmins love Fedora. So we've got that good base and it's smaller than the blue, but it's a really nice solid chunk. The gray there is network install. So I think this is an interesting thing as well. We did not have a network install for Workstation up until Fedora 22 release, and we told people, if you need that, we make that for the server. But we got a lot of people saying, you know, I really want this for Workstation. I use it because I install my office uses this and I have a lab, and so they really wanted to use the network install for that. And so I think that is kind of both shows that Workstation is being used in places more than just people's personal desktops. And it's also a nice response where people said they wanted something and we gave it to them and you could see they're getting it. So that's an interesting thing too. In addition to the additions, we also produce labs and spins. Spins are basically desktop environments meant to showcase that technology. Like the KDE spin is a nice KDE desktop. XFCE is again popular among sysadmins. And so this is including all of those things view there. And you can see that those have a nice big chunk of things there. And also it's kind of pretty impossible to read. I tried to make a pie chart of this, but it got even more impossible. So I think the time sequence, again, this is for the time frame that we've done the Fedora additions. You can see it how that basically the other spins it. So if you're interested in the spins, this one breaks out. This is not a stacked graph. So it's just the individual spins charted there. And a couple of things I think are interesting. KDE is up in Fedora 22 and on. I think that's a lot because we did a redesign. The web team did a redesign page for downloading the spins that make it a lot more prominent and easier for people who want that to find it. And KDE, so all of the desktop spins are up. But KDE in particular, there. So Plasma 5 might be another thing, so people were excited about new school technology in that. And are we one of the first major distributions to have it as well, right? Yeah, so that's faster and first there for us as well. So those are the desktop environments. We also have these labs, which are basically, we separated them out from the spins. By the way, these used to all be spins. We separated it out because some of these things, like KDE, are quite a lot more popular than some of the other things, like robotics, which are a very special purpose. And so we wanted people to be able to find the popular desktops in a way that didn't confuse it with all this, like, here's a bunch of random things. What am I looking for? But things like robotics are maybe not as popular in downloads, but the people who do Fedora Robotics do World Cup robotic soccer. And they often win using Fedora. So I think that's super awesome. We can really appreciate it. And so it's sometimes in downloads doesn't show impact. And I guess one of the, we have a new cinnamon spin. That's pretty cool. But one of the ones, that little black line at the bottom is the security spin here. And so the security lab, I'm going to make terminology, is made for basically doing forensics and penetration testing and things like that. And it is very interesting to use in schools. So people who are interested in that aspect of computer information technology often use that. And I'm pretty sure that that's what the spike is here. Some class that decided, hey, we're going to use Fedora security spin. Yes, Peter? Yeah, I think that's be. Yeah, that's a good question. I don't know that. If anybody has any theories? Yeah, maybe it's the time of year. I do not know. It's a lot easier to upgrade. Yeah, right, yeah, so that is also true. And actually, you can see that. Yeah, why is F23 down? And so go back to here as well. I think Josh has it. I think upgrades are a lot of it. So you can see there is a downward trend in the peaks for the releases there. And that's because upgrading is now using FedUp or the DNF system update. So only people who are new are downloading it. And I would obviously like for that to be going up, too. But I'm not surprised that it's going down. A nice goal would be to have this going up as well as upgrades going up. But yeah. Just to confirm the earlier data of it, we saw it correct that during the Fedora 23 release, the net increase. Yeah, by quite a lot. Yeah, so it is going up. The net number of people connecting is going up overall. And yeah, that's, I don't want to spend too much time. But yeah, so during the Fedora 20 time frame, that was a very successful release. But we didn't see a net increase. We saw we were kind of winning back a lot of these users who had been hanging back on Fedora 14. Suddenly decided, OK, maybe I'll try it again. And then after that, it's when we really see the increase here. So architectures. Basically, this is the picture of the rapid decline of 32-bit until we hit something like a floor here, around 20%. And also, there's arm, a little bit at the top there. That's 32-bit arm. The 64-bit arm revolution is yet to be televised. So yeah, isn't it? Yeah, right, exactly. So I'm not surprised. There are two or three connections there, I assume, from John Masters' house. Yeah, and Peter's as well, yes. Yeah, so I think it's interesting that 32-bit has hit that level there. One of the things for the web design team, I'll change my monitor, I can't show it easily. But one of the things I think might be part of the possibility is, when you go to the download page, there's a banner at the top with a green button that says download that gives you the 64-bit version, which is supposed to be the main version. But if you're like me, sometimes you skim past all that pretty stuff at the top. You're like, OK, let's get to the meat of this page. And when you do that, there's another download section where 32-bit is the top thing there. So I'm going to talk to the website team about doing some A-B testing to see what happens if we put other things at the top of that list and see if those become popular as well. It's an interesting thing. So here, moving on, Fedora Apple. So this is extra packages for Enterprise Linux, a thing we've been doing for a very long time. Rebuilding Fedora packages for REL and for CentOS. And this is something that is immensely popular. I think the graph going up there shows the rise of REL and CentOS more than it shows anything else. But it's really cool that this is popular and the wider community beyond people interested in just the Fedora distribution really benefit from the work we do. And so this is kind of a neat point of collaboration between Fedora and Enterprise-focused distributions as well. I actually expect this number to be going down in the future due to some of the things Redhead is working on to make software available to people. One of the things here is this is a way that people who are users of REL and CentOS get software that is in a nicely convenient package format that they have some trust for, because people have trust for Fedora and Apple. And they feel like this is a great way to get that. And Redhead is doing that in a better way coming up. And so I expect this to actually level off. And if Redhead is successful in what they're doing, they're actually declined. We'll see how that goes. What's that? That's a challenge for Denise. A challenge for Denise, exactly. OK, so if you've seen me give this presentation before, you've probably seen a variety of those kind of numbers. But I'll have a different section. How am I doing for time here? Oh, I got plenty of time. It's awesome. I'm talking fast, so it's working out. I did a director's cut version of this talk, by the way, where it was like an hour and 10 minutes total. And we taped that previously. If you want to go into more depth on some of these things, I've got that, too. So the question here, I often asked, wait, I'll go back before I explain what I think. How big is the Fedora contributor community? People come to me all the time with that. And traditionally, I have made up numbers and told them those numbers. And I thought, that's not really the way to do it. Why don't we do some research? So this does take some explaining here where these numbers come from. We have a thing called Fed Message, which is a message bus. And basically, it is a collection of activity from a lot of different infrastructure and systems in Fedora. Whenever anything happens, they throw a message under the bus. It's like Twitter for the systems that put together the Fedora distribution. And so a lot of times when either automated activity or when you do something as a Fedora contributor, messages generate it. So of those messages, so this is why I have this little star here, there's three key areas that are easily easy to count. That's what I'm measuring here, because that's what you do. So I'm looking for messages that actually have a username attached to them, because I want to actually look at user activity. And of those things, a lot of those are actually stuff that happened about that user or happened maybe as an automated process that's related to that user. So I narrowed it down to things which are actually caused by a user doing something. So you do some activity by yourself, not as an automated process, and a message is generated. And also I narrowed it down to the ones of those where it's a person who's logged in with their Fedora account system account so that I could actually correlate it to a number of things. So that means I'm counting some things that have those properties which are not the entirety of Fedora activity. So that's a big caveat here. And I should also explain the things that I'm counting. Bode is basically the system by which we push out updates for Fedora. And the thing I'm counting here is when somebody gives feedback on an update. So an update is pending for a security problem and somebody goes and votes, oh yeah, this solves my problem or at least it did not break when I installed it. Or they say, that's just terrible. It makes the security problem worse or it bricked my system when I installed it. I count those comments. So that's the number. And so what I'm counting here is the number of people who were in this chart active at all if their name ever showed up, they're counted here. And so that's about 1,000 people there. DistGit is our packaging repository. So this is somebody made a change to a piece of software in Fedora and pushed it. That gets counted there. And the Wiki, you might think, oh yeah, I know what a Wiki is. It's documentation. You would be horribly, horribly wrong. Those of you who have looked at the Fedora Wiki and used it know what I'm talking about. Our Wiki is not a documentation site. It is a whiteboard. And it happens that some people have written documentation on the whiteboard and maybe taken a snapshot of it or something. But it's really an active contributor working space. And I really mean that. The QA team uses it for test matrices. And we use it for drafts of all sorts of things. I counted like 20 different activities that are not documentation. That are the primary things we use the Wiki for. So the important point is this doesn't measure the docs team. This really measures a whole bunch of broad section of contributor stuff we're doing here. So the bottom line there is over 2,000 contributors in all these areas. And there's not as much overlap as I expected. I kind of expected this to be a really like a tightly coupled Venn diagram where most was in the overlap. And that makes me suspect that a lot of the things that we are not counting also don't have a big overlap. So translations, documentation really isn't counted here at all. And then everything the Fedora ambassadors do doesn't necessarily generate a Wiki page at it. So there's probably other bubbles that actually also have their 500s in them that don't overlap on top of this. So plural thousands is the number of contributors. And this is just in 2015. So to narrow that down a little bit more, I basically took the top 10% in each of those areas and did the same thing with that. And so this is basically 300 users who are in the top 10%, which is quite a bit of work in at least one of these areas. And I've got some slides coming up there. I break this down, and I'm going to just skip through those in the interest of time. So if you want to hear more about the breakdown of that, you can either watch the director's cut online or talk to me about it later. I think it's some pretty cool analysis, but it's super geeky and hard to explain. But the thing that came out of that is that this top 10% is about 2 thirds of the work that's measured here, as measured by this, again, easy to measure thing. And that actually is true across all three things that 10% is about 2 thirds, which I think is a pretty good ratio. We've got a good core of people doing things, but it's not like there's a frequently cited thing in social media, which is like the 90-10 rule, where 10% of the people are doing 90% of the work. And we're a lot better than that. We've got a lot broader community that with only 2 thirds. I think it would be nice if we get to the point where that top 10% is only doing 50%. We've got a bigger, a bigger long tail of those casual contributions. And I think one of the things we want to do is make it easier for people to make those lower commitment contributions as well. But I think overall, this shows pretty healthy user numbers. Those nine in the middle are probably the most crazy active people. And I know some of them won't name names, but some of them are definitely in this room. But not everybody. There's people all around the world who are in that. So another question we get deeply related to this talk is, does everybody in Fedora work? How much is this a redhead joint? It is as opposed to being a redhead joint collaboration. And this is just looking at that top 10% because you can easily look and say, OK, these people have redhead.com addresses. And that was a quarter of the people. It turns out that there's a lot of people who are not using the redhead.com email address in Fedora. A lot of people use fedoraproject.org. I actually looked up where that resolved to before I count this down. But a lot of people, I say, are sneakily using other domains. And I need to explain that sneakily. I don't mean it in a bad sense. A lot of these people, and Denise was hinting at earlier that redhead hires a lot of community people to work on the projects full-time because, going back to this, it's hard to be in that number nine if you have another job or doing something else. Like in order to be in the top contributors, it's hard when it's not your full-time job. But it actually turns out we have a lot of people for whom it is not their full-time job. Anyways, people hired in. And a lot of people have established identities in the community already. They just keep using that. Or, like me, they have a whole bunch of filters that they've set up over the years to deal with the stream of email that comes from all of the Fedora mailing lists. And so I was not about to try and port that into my redhead email system. I just keep my email going to where it used to go. So my email address is not redhead. So anyways, that turns out that of these core contributors in the last year, it's really about a third of them are redhead. And we have a huge amount of effort that is put in by not redhead people. And it's really important to recognize that redhead has a gigantic stake here. But Fedora is not just redhead. And again, I think that's a really, really healthy sign, as well. OK, here's the slides. I'm going to skip. Don't they look pretty? See. OK, I'm going to take a drink of water because I've been talking fast. And I have five minutes left to talk about where we're going in the next year. It's awesome. I can do this quickly. First of all, the release train will keep running. We do these two releases every year. It might not be fast from a continuous delivery point of view, but it sure feels fast to all the people working like crazy on it. And I chose a steam engine on purpose because our process for this requires a lot of mechanics, a lot of the time. So it is not exactly a bullet train. But it does keep going along. And we are going to deliver Fedora 24 in June, and then Fedora 25 in early November. And some of the cool new features. One of the things, we are going to have a service. This is something that is an internal feature. But it will not be of any user-facing thing, but it's really important. We're going to have a service by which users right now, contributors can sign up to be the maintainer of an RPM. And now we're going to have the ability to sign up in very much the same way, to be the maintainer of a Docker image. So just like we can deliver RPMs as a primary part of composing Fedora, you'll be able to deliver container images that can be used then to put together parts of the operating system. And so the release engineering team is working crazy to deliver this for Fedora 24. And in Fedora 24, we are not going to be using that to make anything. It's just going to be an available service. And then over the summer or the winter, if you're in Australia, we will try and figure out what the best practices are for that and how that should go. And then by Fedora 25, we expect to be shipping things like cockpit, maybe Kubernetes, as actually containers in Fedora Atomic and things like that. Other things for the desktop, possibly Wayland. If you're a graphics geek, you think that's really cool. If you're a normal user, hopefully you will not notice it at all. And I think that's the goal. Make that transition to a next generation technology stack as seamless as possible without making our upwards going green line all of a sudden go, whoa, what's this new change? While still giving people the benefits of that, we're going to make that decision pretty soon about whether we're ready for it in Fedora 24. If not, Fedora 25 is looking pretty good for that. We're getting real close. Another neat thing will be upgrades in the GUI. So right now, like I said, you don't need to download an ISO to do an upgrade anymore. But you do have to drop to the command line, where Richard Hughes is working on this awesome thing where it'll just be like my phone has been sitting here this whole week telling me I have a new Android update. It will do the same thing. Not with Android though, with Fedora. And then, of course, certain things will have new versions of all the stuff, because that's what we do. And Fedora 25, we'll see where that goes. Another important theme this year, marketing. This is the market in Brno. I don't know what picture you put to represent marketing. There's a bunch of diagram-y stuff that looks silly. These releases we're putting together are awesome, and we need to get more people using them. We want that green to keep going up, and we're going to need to reach out to places that are not traditional Fedora users in order to do that. It's encouraging that we have an upswing. We need to make sure it keeps going. And the way we'll do that is by looking at people who are not using Fedora, asking them why they're not, what their problems are, what no operating system is solving for them, what no Linux is solving for them, and then coming back, delivering those features, and then going back to them and saying, here, we made something you wanted. They say, whoa, that's what I wanted. And that's the thing, that's the cycle. In particular this year, Fedora Workstation, I said, was targeted at developers. We are talking about this is not a done deal, but one of the ideas we have is that developers is still a really broad audience. Maybe for this year, let's bite off a subset, possibly Python developers. Fedora loves Python. I didn't have to make this graphic. It already exists, because we love Python that much. And so this is a narrow subset of developers we already really have a good relationship with. And so going out to the Python developers, like Ian who's nodding here in the front, and solving their problems. And since we're short for time, I won't go into all their problems, but they have some. Out of time, another initiative in marketing, Fedora University Outreach, Fedora Hubs, which will be basically bringing us out of, yep, that's right. This is basically, there's a whole bunch of stuff where Fedora activity going on all the time. We have like thousands of IRC meetings every day, over every day, every year, over three IRC meetings every day on average. And all that stuff is buried activity, mailing lists are something that we're really used to as open source contributors in Fedora, but most young people today think email is obsolete and mailing lists. Oh my God, what a terrible thing to have to deal with to do anything and get anything done. So Fedora Hubs gives us a new interface.