 Hi everybody, I am Matthew Miller, the Fedora project leader, and this is part two of my state of Fedora 2016 presentation at DevCon special directors cut edition made for made for the internet direct to you. And so the first part of it, I talked about some of the numbers about the download statistics and where we were in 2015. Here is where I'm going to look ahead with basically no numbers into the future because the future, you know, don't want to put numbers on things so then they'll be wrong. Except for Fedora 24 and Fedora 25, I'll put numbers, those two numbers out there. Of course, in this year, the release train is going to keep going. Fedora has a schedule where we put out two releases a year and our target is generally May and late October, which is a code word for November. And this year, it looks like it's going to definitely be June and November. That's a lot because our release process targets those dates, but is a blend of feature based releases and time based releases because we basically integrate thousands of pieces of software from all sorts of upstreams to all sorts of different things. And we try to get them aligned at those times and nicely Q8 and tested so that we can deliver them to you, the users in a way that is not broken. And so if we would try and do that with a hard schedule, we would always deliver something that's broken. And if we would try to do that with something where we want to get all the features lined up without trying to impose a schedule on it, we would end up giving you a release once every eight or nine years. So instead we have this hybrid approach and although it kind of makes people frustrated because it's hard to predict the actual date and it always feels like we're slipping, it's actually not failure. That's the process and it's really worked really well for us. So I can tell you that we are going to have two releases this year and talk a little bit about the Fedora 24 features. Again, I was talking about marketing and the first part of this about how release driven features are kind of better historical marketing. But one of the things is we're getting to a place where it's really pretty good and so a lot of the changes we're making are under the hood. So in this new release there will be new versions of software. You'll have Darktable 2.0. You'll have whatever the next version of Ruby is and all of your libraries and all those things are going to be updated. So that's one of the things you can count on in a new release. And we also have some new additional features. One of the cool things is going to be in our infrastructure. If you're here at DevConf you know that about half of the talks are about containers. It is a containerized world that we are moving into and Fedora is preparing to be ready for that. And so one of the features we have is a layered image build service. And so this is not an end user facing feature yet. But this is a thing where right now as a Fedora contributor you can take ownership of a package. You say I want to maintain some library or a game or a utility and say I'm going to provide that in the Fedora infrastructure and that's you as an end user that can do DNF install or go into GNOME software and search for it and then it's just installable. We don't currently have that for containers. And so we are going to starting this spring. You'll be able to be a container package and you'll be able to say I would like to contribute to the web server container that Fedora will offer. And you can basically using sort of the same infrastructure we use for packages produce these layered images that are basically docker images at this point that we can then actually use to compose parts of the OS. And so this is going to be important for Fedora server. If you're familiar with Fedora server in addition to being a platform for building whatever your services on to the key feature of Fedora server is roles which are basically easy push button thing I would like to turn this server into a database server. Boom you push a button and then you have a database server and with very little it basically does best practices configuration for you. And so those can be delivered as containers rather than being delivered piecemeal on your system. Another thing in atomic which is basically an optimized version of Fedora just for running containers. Right now we're delivering the atomic host but we're not delivering anything that you could actually use it for. So having the layered image built service will let us start building the things you can do with atomic which will obviously make it a lot more useful because right now you have the base and we expect you to build or find what you want to do with it. And so for Fedora 24 that's kind of an inward facing thing that is going to be available. If that's exciting to you come join and become a contributor to that. If you're waiting for it to happen and be something that you can use Fedora 25 in the fall will have some of those things. So that's exciting. Switching to Fedora workstation if you are using on your laptop or desktop one of the big things that's going to happen is that you'll be able to upgrade to this release using the software updater. So GNOME software which you can use right now to update to security fixes release. You'll be able to say I want to go from Fedora 23 to Fedora 24 just like if you have a phone running Android and it says there's a new release available it will do the same thing. You click upgrade and you'll be right to the next release. And I think this is going to be really nice for people who keep saying it's really a pain. If it were on such a life cycle after reinstall it's why don't you do a rolling release instead? Or why don't you have a five year life cycle? Instead we're going to just make it push button to go to the next release and so it will be a lot less of a hassle to do that. That's kind of exciting for that. Another thing that's somewhat related to that if you are not upgrading if you're a new installer, a new user and say you know you are stuck on some operating system that is not Linux. It's actually kind of hard to figure out how to take our download and make it into something you can install from. We've got some complicated technical instructions for doing that. What we're doing now is actually switching the main download. So instead of downloading just an ISO image and you have to figure out how to burn that to a USB stick, you will download a program for Windows or Mac or Linux that will actually write to the USB stick which we hope will make it a lot easier for people who are switching from proprietary operating systems to come to Linux. So that's going to be kind of a nice feature as well. There are some other under the hood things. If you are a graphics technology geek you might be excited for Wayland. Wayland replaces the traditional X11 graphics environment in Linux. The ideal here is that if you are not one of those graphics geeks you will not notice or care that this has happened. You will just get a more secure desktop. If you ever played videos in Linux and you notice that there's Flickr and it doesn't sync, Wayland will fix that. Games will be better. It's a better world and it allows us to do better sandboxing to have secure applications at running containers. Again, containers, containers, containers. But if that's delivered properly you won't tell that it will just be better. It won't be like, oh my goodness, we are doing bleeding edge technology that doesn't work. So we are going to switch that. That's available now in Fedora as non-default and we are going to make a decision about whether it's ready to be the default for Fedora 24. And if it's not ready we will wait until it feels like a painless switch for most of our users because we have had some great releases and getting that user base going up and we don't want to break it and have another Fedora Dark Ages. So there's that. And again, like I said, new versions of all the stuff. Fedora 25 in the fall. We'll see what's coming up there. We don't want to look too far into the hazy crystal ball. But like I said, layered images as actually part of the operating system will be part of that. Another one, the atomic project wants to deliver cockpit as part of a container. Cockpit is a really cool GUI for managing system, managing servers. And it's sort of like the new, the web version of logging into the console with a lot of like monitoring and the ability to actually manage Docker containers from the console, access to logs. And it's a multi-host thing for, you know, something, if you want to go up, you know, two thousands of hosts or maybe even hundreds, you want to get something like Spacewalk, one of the Red Hat Satellite, those kind of products. But in a smaller environment, basically the idea is this replaces logging into the console. Although I should put that in the caveat. It doesn't take it away. In fact, there actually is in your web GUI, a nice HTML5 command line prompt right there, which is pretty slick. I demoed this at Lisa, the system administrator's conference. And it was bringing all the system in to our app. It was nice. So that's the release train. This slide's the market here in Brno because I don't know what to show in a slide for marketing. I didn't want to just put a bunch of words up there. So one of the things we want to do this year is increase our marketing and really have a marketing make and effect on Fedora. We've got that upward trend in the graphs and we need to keep that going. And we'd like it, you know, to be like in these hockey stick curves where we exploded the entire world because we've taken over with Fedora. In order to do that, we really need to work on making our marketing work and make it more than just we put out a press release when there's a new release telling everybody that they've got a new version of Ruby because that may be exciting to a few people, but it is not a compelling story. So we need to figure out how to do this in a more effective way. And particularly, marketing is more than just sending out announcements and more than making swag and collateral. It is kind of about going out to users and finding their problems, bringing them back, solving their problems and then coming back and saying, look, we solved your problem. And so that's one of the things we need to do, including reaching out to people who are not Fedora users and say, what problems do you have that no one's solving? Let's solve them. Particularly this year, Fedora loves Python. This is kind of a new plan we're putting together here. Fedora Workstation has always had one of the primary targets to be developers. We're going to make this a great workstation for, that's what we call it, workstation instead of desktop or tabletop or whatever we could have done. That's an audience we really want to appeal to. It doesn't mean that it is not an awesome thing if you're not a software developer, but the idea is we're going to make it great for software developers and then grow out from there. And this year we want to specifically target. And again, this is a work in progress, so help wanted with this, but we are thinking that we want to make Python developers in particular a target that we're going to reach out to because there's a lot of love for Python already in Fedora. I didn't have to make this graphic. We had a Fedora loves Python graphic. And so we're going to kind of bite off something. Developers is still a really big market and we wanted to bite off something that we could actually, you know, chew. I don't know if we actually want to. Anyways, we're not going to bite people. It's a terrible metaphor. And there are some specific things we can do in Python. Python is talking about providing an ABI so that if you have binary Python libraries that they will work from system to system. We can try and work and make sure that Fedora provides that. And we can find... Basically, that's the thing that I know about, but what we need to do is go talk to a bunch of Python developers and say, what do you need? And so if you are a Python developer, come talk to us. What do you need in your wildest dreams of an operating system for doing your Python development on? What would that look like? And we'll see what we can do to make that be things that we're delivering to you in future Fedora. So that's part of the plan. Another thing is Fedora Hub. No, wrong slide. That's okay. I'll do the slide. You don't need to read the details here. This is just to add something on the screen other than me giving you bullet points of what we're doing. One of the initiatives we want to do as well is university outreach. So this is kind of our high-level plan for what kind of things we're going to do to talk to universities and get more young people involved. Because even if developers are our target, we need to get the future generation of developers so that when they go into companies and they say, you know what I want to use? I want to use Fedora because that's where all the cool stuff is. That's what I'm familiar with. That's what's useful to me. So we're going to ramp up in this next year our outreach to university students. And we're going to try and do that with meetups and smaller group activities rather than going to traditional Linux conferences and things where we kind of hit an already exposed enthusiast audience who already kind of have an idea of what Fedora is about. We can tell some of those people, you're wrong. Fedora is better than you think. And that's got some potential for growth. But we're going to have even better growth by going outside of that to people we haven't talked to before. So that's part of that. This is the slide about Fedora Hubs. It is again something that is basically to be on the page because you don't want to be just reading what I'm saying. And Fedora Hubs, I talked earlier about Fedora activity being an iceberg in that if you look at the Wiki, you see maybe some edits going on. You look at the web page. You know, we update the download Fedora web page every six months when there's a new release. It doesn't really look like there's a lot going on because on the internet today, the web is for the vast majority of people even highly technical people. That is the internet. And so when you look at Fedora for us, the internet is email, which as we learned from slash dot a decade ago is for old people. But email is our activity and IRC. We have all those IRC meetings. You have three meetings a day in IRC where we do all this really hard work. But none of it is visible to anybody. And also, although those tools are great and I love email and I love mailing lists and I love IRC and maybe some of you do too because you are also crazy, there are ways that we could have an interface that is better for those things. And so our design team has been putting together mock-ups for this thing called Fedora Hubs, which to my mind is sort of a social network for Fedora development, although there's kind of a trend that I hope has died for a social network of everything. So it's not quite like that. Maybe it's more like a portal, and terrible, but basically getting us onto the web, that's really what it's all about. And so the design team has been working on this over the last year, and this year the infrastructure team and Paul Fields, who's the manager for that team inside of Red Hat, has committed that they're going to put resources into this. And so this will be the year when we have Fedora Hubs and I can stop trying to explain to you what I'm talking about and I'll be able to show it to you. So that's the excitement there. So Project Atomic. Again, containers, containers, containers. And I talked earlier about how we have this atomic version of the Fedora Cloud edition. What we're going to do this year is actually switch it so that Project Atomic is a top-level edition. So you have Fedora Workstation, Fedora Server, and Fedora Atomic as the top-level things. That doesn't mean that cloud isn't important. It means that cloud, as it is, is kind of a deployment environment and underlying technology for things and it's not really a target audience. When we first invented that as one of our top-level editions, we were definitely thinking about it as a target audience and there is people who are doing cloud computing and infrastructure for applications in the cloud. But with what we were providing with the Fedora Cloud base image, we never really had a compelling story. And I used to work on Fedora Cloud as my primary job in Fedora because this is the director's cut version of this. I'm going to go into some detail on this. I would go out to talk to people and say to the marketing, what are your problems? What do you want from a cloud operating system? And they would routinely say to me, well, I want it to have a basic little tiny core that I don't have to think about and then I want to have stacks that get lumped on top of that. And it'd be awesome if you could provide for me different Python versions, different Ruby versions of Java, all of these things, there's different stacks that I can plug in on top of that and put my applications on top of that. And so with Fedora Cloud, it never, the cloud base edition as we had it, we kind of got the minimal thing going on, basically a minimal install of Fedora with cloud in it and some utilities on top of it. But we never really got those pluggable software stacks happening in a way that made a compelling story. And so long came this idea that Colin Wolters had actually presented first at DevConf here two years ago, I think, amazing of using an OS tree, a basic technology to make that minimal core image in sort of a way that can be updated basically in the same way you update a Git tree, you check out a new version of it rather than piecemeal updates. And that kind of is a really good way to make this core foundation. And that grew into this whole atomic host idea in Project Atomic where we'd have this minimal, small optimized OS for containers and then use container technology to deliver those pluggable environments that people want. So Project Atomic is basically an answer to the thing that people said that they want. And so we are doing that in Fedora, the bleeding edge, the leading edge, I try to say voidsaying bleeding edge but we're kind of in the bleeding start part with Atomic but it's kind of get less bloody and more to just the leading part where we are in the forefront of doing this in Fedora. So right now in Fedora, we have a two week release version of this so every two weeks a new image is automatically generated and tested and available for download in Fedora. So that's already happening but right now it's kind of a side thing. So we're going to put that front and center I guess as we go from bleeding into actually leading it becomes useful for people when you have these things to run on top of it available. It seems like a really good time to make this be our front and center focus. And that's not all because that approach of having a base image and then layered environments to go on top of that and Atomic in specific containers like that's not just this cute thing for the cloud this is really the future of the operating system and you don't have to believe me you don't have to take my word for it again look at this conference that's half container stuff go to any Linux conference LinuxCon is now LinuxCon slash container con and all of these things this is what's happening it's the way the innovation is happening in the OS and Fedora has a responsibility and a job and a charter like we're supposed to be in the forefront of this and this is the way we're going to do it and if we don't we are going to be left behind as an operating system so we need to get on board with this and at the same time Atomic the project is becoming increasingly connected to OpenShift which is a completely open source platform as a service that is also out of Red Hat and that kind of provides developer infrastructure around the Atomic container framework and that is going to be increasingly tied together and I think as Fedora we also need to increase our relationship with OpenShift because those lines are going to be blurred and soon the platform as a service operating system and that's a really cool exciting future so we need to be part of that and Lego bricks as part of that you've seen Lego in my slides before if you've been following the Fedora Next things this is again part of going to that operating system future we need to work on actually modularizing what we have in Fedora so that instead of a bag of bricks we could provide different solutions to people that are actually assembled at a higher level let some bricks I don't have a slide for it but bricks that are put together into preassembled sets that have been tested as a set and we know this is a good spaceship I don't know if you did this as a kid you take your spaceship and you throw it against the walls see how well you built it that's what we're going to be doing with parts of the operating system so we will see that when we have an update to something you know that some kid in Fedora has thrown that against the wall and it did not break awesome metaphor I'm going to go with this one we're going to really work on continuous integration and continuous testing infrastructure so that we can deliver these parts in a way that is not scary because one of the scary things about this is as a sys admin when we went to making these nice Lego bricks that was a big improvement because the previous state of software and open source software in particular and proprietary software had its own set of problems was not that you had Lego bricks nice packages but that somebody gave you raw molten plastic sometimes still hot and said here here's your software and then that is really a nightmare for a sys admin to deal with that and so packages and the packaging guidelines we have in Fedora distribution and other distributions came along and really solved these problems for sys admins and so that's a great thing but it doesn't scale and it doesn't work in the world today developers I'm a sys admin so I'm going to say it this way as evil as they are those developers run the world now and management listens to them not the sys admins this is a battle we've lost and it's okay because it's good for the users because it's about the applications it's all there for the users and the applications are there for the users and the developers make the applications so it's okay we're losing this battle but we don't want to go back to that crazy molten plastic world we need to go to a new world where these containers are produced and tested automatically and we actually have a better solution than having little bits of blocks so we're going to be working on that this year in Fedora and I think I don't think that by the end of this year I will have a complete solution and be like look we containerized your OS it's amazing because we're not going to break it as we do it we might shake it up a little bit but we're going to make sure that we don't undo all the progress we have in Fedora growth by scaring people too much with too much change but this is going to be the year of show and tell or more show less tell I've been talking about this for a while this year we're going to have prototypes we're going to have people getting involved people showing up you know shovels and hammers we're going to make this and I think that's going to be an exciting exciting future for Fedora thus ends the crystal ball now is the time for the thanks I'm Matthew Miller the Fedora project leader thanks very much for the earlier part of this presentation to Steven Smugin and Ralph Bean who did incredible work pulling together statistics for me and especially as I was working with them and hacking at all hours to make these graphs and having a crazy idea about what if I did it this way a lot of back and forth and so they really also worked at all hours to support me so thank you very much for that thanks to the entire Fedora community because like I said you know thousands of people this happened I'm standing up here talking to a video camera but I'm figurehead for a lot of this and it's really the community that does it thanks to Remy who's here off camera who gave a presentation of this at Fosdown which is very well received and I was able to use that to sneakily edit my comments to address comments as I talked so that was pretty awesome and Remy is going to be doing, Remy is the community lead for Fedora and is going to be working on a lot of the awesomeness that I'm going to need to support the future of this and finally add a lot of statistics you too can become a statistic if you're not a Fedora user fix that right now go to GetFedora.org and join in and if you are a Fedora user or if you are just so inspired that you want to become a contributor right away that's awesome what can I do for Fedora.org will tell you how you can plug in and there we go thank you very much