 Well, thank you. As Frédéric said, I'm the project leader of Krita and I've been working on Krita since 2003. That's quite a long time. We've got a lot of history there. Part of the history of Krita is quite interesting because it sort of does show the way a free software project can grow and mature and change over time. So, I wanted to apologize. I didn't do a lot of work on the slides because it's been quite a hard time for the past three weeks. So, history. How many people knew about KIP? K-I-N-P. This was, I probably have told you who are there, I guess. Way back in the prehistory when people were still innocent and GBK was only the give-it-all kit, people started working on this, this weird idea. Let's have an integrated, sophisticated Linux desktop system where all applications use the same kind of conventions for Google. So, what we had was KDE, Qt, and we had this painting application called GIMP, which stuck out and didn't look very good on the KDE desktop. So, whatever, some of the broad patch and patch may give you Qt and all the conventions. That led to a huge row. This patch should not have been written. It should not have been made public. People have talked about it. In answer, the response from KDE was, let's do our own GIMP clone. Let's do an application. Let's give it a really original name. We call it K-ImageShop because that's for sure not going to give any problems. Our name is still called K-ImageShop, but that's the last random of that name because, well, lawyers happen. So, we had to rename the next name, Crane. That's nice. It started with a K. It has something to do with painting with the lawyers. Finally, we renamed to Krita. A couple of years later, the Krita.org domain name became available after the music group of the same name stopped. That name sort of stuck. I'm still getting requests for renaming because it's so bloody hard to pronounce. It starts with a K. I don't think it's that hard and it even makes a good furb, right? I'm going to totally critter this image. I don't know why I don't hear that, but I still hear people talking about Photoshopping image. Back in those days, the few three days when we were finally working on the release of Krita, the project was almost dead and that's quite a common occurrence because you start enthusiastically with your free software project and then you realize that, oh my god, it's going to be so much work. It's going to be so much more work than I'd ever imagined I'm going to stop. Actually, that happened to me as well. In 2003, I got a graphic template for my birthday because I wanted to draw a map for a fantasy novel I was writing and I was trying to do that with GIMP and I didn't understand GIMP. That's my fault, not GIMP's fault. So I started my own painting application in Python and Qt because I'd written a book about Python and Qt and after two months I realized that I totally didn't have a clue. I could write an application that would show a canvas, it would show a manual bar, a tool bar, a status bar and it couldn't draw a line. So I had to look for some project that would make a good start for me. At that point, Krita had three rewrites of its internals in six years. It hadn't had a release. It had three rewrites. It could load an image, it could show an image, but it could not draw a line. And it was written in C++ which I didn't know. So I started working on it and by then I sort of didn't mind admitting my ignorance. So I started blogging about it. Look, I can make a red square and if I press really hard with my stylus on my tablet, the red square gets bigger. That was quite, quite something. A little later I was blogging, I can make a line but the lines are horrible. They are getting banded. It's like this. If you're making a drawing application, what you're actually doing is you draw footsteps off your brush with my right hand. And if you make enough of those, you will get a line. However, drawing those steps can be quite time consuming. So if with my first painting tool in Krita, you wanted to draw, say, a circle, it would look like this because the system would draw all the events between them here and it couldn't interpolate. And that's something that every painting application struggles with. So what I did was I blogged about it. I am trying three different ways of making a nice round circle in Krita by drawing with free-hand. And I'm not succeeding. What am I doing wrong? What happened is that suddenly we had four more contributors. There was this guy from England, Adrian Page, and he took my line drawing code and fixed it. There was this guy from Denmark and he took the rendering code and rewrote it. And there was this guy from Germany and he started implementing some of the doctors and there was this guy from France and he started implementing filters. And I was like, this is fun. And we had a lot of fun. We finally made a release. The guy from Denmark called me a slave driver and he was right because we were going for this release. It was going to be the first release. Three years of work. Just when we were going into beta, one day before we released the beta, we had this new back-hand rewritten. That was the fourth time we rewrote the back-hand. And finally we had infinite layers. Because there are two kinds of painting applications. One where every layer fits size and if you move the layer, you still can't draw outside it. The other one is where you can draw outside the original size of the layer. Krita started out as a game clone so it was the first time. And then we rewrote it to be the second time. Release Krita 1.4. We got reviews in the Linux press because this was still old days, 2004, 2005, when the Linux desktop was hot and we were still trying to create a complete replacement for a proprietary system. That's something that actually for me changed in a way because I'm not trying to totally replace an application on my application. I'm trying to replace something at its own thing. But in those days we were going to do everything that Photoshop could do and we were going to do it on Linux and we were going to do it integrated with KDE. That was Krita 1.4. It got reviewed. The reviews were really positive because it was a new application and it could do some stuff, nothing else could do. And then we released Krita 1.5, which had CNYK support. Which is, and at least in the way we implemented it back then, totally useless for actual users. Digital artists don't paint in CNYK. They paint in RGB and convert to CNYK, clean up and publish. We didn't support that. But we had this tick box, CNYK, OpenGL canvas, filter layers, non-destructive editing. We were complete except that we still didn't have users. We didn't understand why we didn't have users. So we just went on. We started working on Krita 1.6 and at the same time we started porting to Qt 4. By the time we released Krita 1.6 we had a user. He was a guy from Finland and he did some really nice work in Krita 1.6 and he had some really, really interesting comments about usability and performance and stability especially. He said if it doesn't get any more stable, I really will have to stop using it. In that case you can either say, okay, stop using it and start using something else or start listening. This was more or less when we were beginning to realize that what we were doing was not developing software. We were not developing application. We were having fun writing code. The code mostly compiled. If you would start the application something would happen but it was not an application for users. In my opinion, application that's not written for users is like the sound a tree makes when nobody is listening. It's screaming in space. It's just not useful. But at that time we were also having other problems. The port of Qt 4 was horrible. We were currently porting to Qt 5 and that's also horrible but not as horrible as the port of Qt 4 was. That was really horrible because we were really done. We were doing a complete rewrite of Krita and the other applications in the street. A complete re-architecturing during the port. That's the most crazy thing you can do. Don't do that. We did it. It took years before we arrived at Krita 2.0 where it sucked. You could not make a selection and cut pixels. Krita 2.1 sucked as well. It sucked a little less, but only a little less and it was totally unusable. Krita 2.0 was just as bad. That was the moment when a French guy, David Rois, he started moving from proprietary software to free software. He is an idealist and he wanted to do this because he'd noticed that when he bought a new computer which came with a new version of Windows, he needed a new license for all of this growing software. At that time, he was proud of Photoshop, Coral Painter and a bunch of other applications. He just couldn't afford replacing all of that. He said to himself, I'm going to check what this free software community is giving us. It gave him the software that he could use professionally to do his work, not without limitations, not without loss of productivity, but using GIMP and MyPaint, he could do his job. We just had decided, as Krita developers, that we couldn't be everything to everyone. We were not going to do a complete Photoshop clone. We were going to do an application for painters, for illustrators, for comic book artists, for native pages, lecture pages. So we were doing an application that was free software and that was targeted at this guy that we could not use our application. What do you do if that happens? If you learn that your application is so bad that your target audience cannot use it. You can stop coding or you can start coding even more. That's where we figured out that we wanted to continue. We had this vision. We are going to make a really great application for real artists all over the world. That takes time, more time than we actually had as spare time developers. A bunch of sum of code students, three months of focused work on improving performance, but some things just take a whole lot of time. At this point, Lukas Trudy is from Slovakia. He was almost finished university and he had a couple of months of spare time, which in his country apparently is being used to earn some money before you start writing your thesis. He was also going to do his thesis on Krita brush engines, but he had this spare time and he wanted to earn money. Totally fair. Guy has to eat. He has to pay rent for his apartment. If I remember correctly, his apartment went for 90 euros a month and he was worried that we could not collect enough money to help him pay for his apartment. We did our first fundraiser. It was quite a primitive affair. I took all the money in my own bank account, which is not a good idea, and made for really horrible administration headaches, but we got the money and we paid Lukas and Lukas spent months and months and months fixing all those bugs, all that instability, all those performance issues. And then we released Krita 2.3 and David Rieva said, this is a very nice illustration application. It has potential. It's not suitable for professional use because the workflow sucks. With you guys, you have made a nice application for editors to use. Say if they want to paint some manga. Okay. We improved, not enough, but we did improve. So I shouldn't do that. I'm used to talking with my hands. We were happy. We had proven two things. Krita was improving and funding development does not kill your project because that was a huge worry back then. I remember discussions on the KDEV many missed about how for me destroyed some KDEV projects and fears that that would happen with Krita as well. The idea being, if you're a volunteer developer and you're having fun hacking and someone else is getting paid for hacking, then you will feel bad because you're not getting any money. That's totally logical, but apparently it's a fallacy because it didn't work out that way. We got more money than we knew what to do with actually because there was this, this, this, this, well, of guy who liked Lucasworks so much that he sponsored Lucas for another period. This full-time work really made a difference because it takes the application to another level. So when we released Krita 2.4, David Rivois switched. We had reached our goal, maybe even happy. And we knew that by now Krita was good enough for other people. Linux using illustrators all over the world now could use Krita to paint. That was a huge success. I know several things happened together. I started the company together with Ingo Alin who was also a KDE member and another guy and Dimitri was finishing at university and I didn't want to lose his contributions. So we started looking into ways of accelerating even more. And the thing is, it's a lot like running. Running is the controlled form of falling. If you go faster, you won't fall. So we have been going faster and faster ever since. So if Krita is good on Linux, why shouldn't Krita be good on Windows either? Turns out that porting a KDE application to Windows, if you've got the target of having a million downloads, is hard. The KDE four libraries were not suitable for releasing software on Windows. I had to strip them down a lot. Remove things like Debus. Because people with oversensitive virus scanners got really frightened about Debus. Fortunately, we managed to get some funding. Ingo funded our first effort of porting the whole KDE suite to Windows. And in its wake, it funded creating a special tablet version of Krita and then a two-in-one version of Krita called Krita Gemini, which meant that with this support we managed to release a good application on Windows. That took more than a year of full-time work, which is not cheap. And we really needed that. The thing is, now that we're on Windows, we're getting about 30,000 downloads a month. And that's about a thousand times what I see happening with increase in Linux users. So that means a huge new pool of people who can join the community. And I'm really happy to say that by now we have actually got Windows only developer, a guy who only uses Windows. No, I'm wrong. We've got two. They're both trying to fix the same bug on the Surface Pro with stylus. With this huge growth comes a community. And community is an overused word. What is a community? It's not just the developers. I consider everyone who uses Krita to be part of the family. And I try to foster that feeling among the community using social media of course. But also something really simple. If someone reports a bug for Krita, I say thank you. Thank you for your report. I notice that almost no developer does that. But it's amazingly powerful because only once you've said to this guy who comes to a bugzilla and has taken the effort to register the bug and maybe even make a video of the bug. And even if I can't reproduce it, which happens quite often, of course. And then he's like, oh my God, Krita ate my data and ate my image. And it totally sucks. Oh, what am I going to do now? I say thank you for your report. I'm sorry about your data loss. I will try to help you recover your file if the file is broken. But if you say thank you, you've diffused that tense situation. And that's just one of the things where you can improve your interaction with your users and avoid all those nasty flame fest that happen all too often. Of course now that Krita is getting really popular, I'm also getting really weird reports and really weird mails. Albert, as you can see, who doesn't do anything on Krita, a guy who is working on Kaxi, Adam Pick, and I both got the same mail. Well, I've got a tweet, Adam got a mail and I got a forum message with the same text. I want to do custom development on Krita. I have $150. So I finally managed to compact the guy. And what he wants is re-licensed Krita to a commercial-closed proprietary license, add a license manager, rename the application, re-brand it totally, make sure that there is an undo button in the toolbar and the selection outline is no longer marching ends but solid line. And he wants to pay me $150 for that. I'm doing a lot for money. I really like money. Every cent that the Krita Foundation gets helps from development sprints, fun stuff. But I'm not doing that. So how am I funding Krita development? Krita development currently is funded in a lot of different ways. And I'm going to tell you my dirty secret. I copy everything Tom Rosendahl does for the Blender Foundation. He is making a free software product that kind of affords to pay a dozen developers. And that's just awesome. So what I'm doing is we are selling DVDs, training DVDs. We've currently got two DVDs and there's the third one coming up, at least going to be created, called Secrets of Krita. And it's going to be really exciting. We are selling... the guy running over there is going to make it. That's an investment, of course, but you get it back. We are still selling about a dozen of the Muses DVDs a month. We are selling Krita on Steam. The advantage for the people who get Krita on Windows on Steam is they get automatic updates. There's also a version. They also get the Gemini version, but it's a bit of a problem because that version is broken. So the latest version of Krita doesn't have the Gemini bit, but they also get the latest version of Krita. Automatic updates, Cloud Gemini. That's bringing you money about between 500,000 euros a month, which is amazing. We've got a Krita development fund where people can subscribe and give us a monthly donation. That works out well. We are getting lots of individual donations as well. People who use Krita and decide, okay, I made my deadline. I earned some money. I'm going to make a lot of money and then, of course, the Kickstarter campaigns that everyone knows about because we are really public about those. The last one was a big success. I'm actually considering doing a second one this year for OSX development because I'm estimating that we have Krita on OSX, but it sucks. It's not suitable for professional use. It will take me at least four months of full-time work, really nasty work that nobody wants to get there. Right now, I make those OSX builds just to make sure it stays buildable. So, the points where I'm trying to get to is, as a free software community, the free software community that KES sometimes lives in a bit of a bubble. We read free software blogs, free software websites, Linux websites and we keep up with all that news. There is such a big world outside. That's what made me so, so, so incredibly happy when we got a review in ImagineFix. It's a paper magazine, though still exists, for artists, mostly digital artists through traditional artists who paint fantasy images. Now, there's this whole cultural brain thing that I had to sort the shake off. When I was 18, well 17, and needed to leave high school and go to university, I was sort of, well, I could go to university, or I could go to an art school and do something with my sculpture. But, if you become an artist, you are going to start, right? Art is frivolous, doesn't bring money and there are way too many people doing art anyway. So, come to university and do something useful, like studying the languages of the hill tribes of eastern Nepal. Where's the three film languages? Lots of interesting features. I don't regret my choice. But, when I started working on Krita, I still had the same attitude. Art is frivolous, it's a fun bit. The serious parts of K-office were, of course, the word processor and the spreadsheet, because those are serious applications for serious people earning serious money. Well, artists are earning serious money. I don't know how many artists are being employed by the Game of Thrones, the television series, but it must be a lot. And all of them should use Krita for their concept art, of course. Same goes for all those Hollywood movies. And I happen to know that for some of those Hollywood movies Krita has been used, both for concept art and for creating the Matthew paintings, those painted backgrounds. If you watch a movie, check whether the clouds move. If the clouds don't move, it's a painting. And again, there are a lot of them because it's way too expensive to model a whole world to 3D. Or to go on location. So, for instance, for that Disney movie about the big white guy, I've forgotten the title actually because I haven't seen it. A little big hero, that's it. Krita was used for G.I. Joe. And I'm actively trying to get Krita used by more studios, which is a hard task. You would think it would be easy because a movie studio, a game studio, they are mostly using Linux. They are using CentOS because they are too cheap to pay for Red Hat. They are using the Linux applications from Autodesk. And Autodesk is, as far as I know, too cheap to pay for the cube license. All those applications are using Qt. An application like Mario, which is a 3D painting application, uses PySight because they are too cheap to use for PyQt. So, there is this, and that's probably the reason they are not paying for Krita yet. They really want an application like Krita, a 3D painting application that runs on Linux and is fast enough and has all the features they need. So, I'm pushing in that area as well. And then some unexpected things happen, like last year the first book on Krita was published in Japanese. I was already going to order it when the publisher sent me a copy. It's got lots of cute manga girls in it. So, I hinted in the talk description for the program that I was going to talk a bit about plans, and I've got 20 seconds, so that's going to work. We are going to maintain the Qt 4 version until the end of the year, doing monthly releases with features. We are working on animation support, some of code project, but it's also something that the foundation is funding. I'm working on the Qt 5 port. I'm struggling with the open gel part. I'm getting black screens and nothing but black screens, and I don't know why yet. But Krita actually uses some weird open gel stuff like gel fencing, which apparently nobody else uses. I don't know. It can't get it to work. Okay, well, I'll show you the code. I didn't write that one. At least one Kickstarter per year. Maybe two, because O6 is a special case. O6 users don't want to pay for the general Kickstarter because they can't use the result anyway and they are right. We are going to do the DVD and there's going to be a book published in English. It's written already. It's in the editing phase and there's a publisher. And then next year, well, I guess we'll try to do even more. Thank you. Thank you so much. Any questions? What are your plans for Qt 5 vendors under the EOS program? I don't know. I hope I don't know. I hope that it's still possible to build Qt 5 without Venus. Yes. And if so, I will use my own homegrown Qt. Just like I do already, because the Qt we use now on Windows and O6 doesn't have Divus and doesn't have OpenSSL and a lot of the other things that we don't really need. We're still packaging way too much stuff. What do you use Divus for? We don't. They've had it out and that still works. The thing about Divus is well, on Linux we've got this beautiful Divus interface you can use to create new Windows open files or stuff like that. That's a straightforward of the Decop stuff that we have from Chaos 1.0 or something, right? Chaos applications, if you're pouring away from Chaos applications, how am I going to make it? Yeah, we are doing that. So on Windows apparently that triggers alarms with some users. The thing is if you're deploying a Windows application to a corporate environment you can do stuff like that. If you're making a Windows application available for download for Windows users all over the world and anything that opens up software is going to be a problem. I don't know why, but it's true. I understand that, but if you don't need it then I don't get the issue what you have in there. Although we need this kind of a Divus interface. KDNX4 I supported it until the end of the year. I needed it until the end of the year. And I don't know about frameworks that can be completed without Divus. I don't know yet. Well, you might have to skip a few frameworks so the whole question is whether you need them. Right now we need even KDNX4 support because we haven't been able to cut out all of that yet. First, I want to fix the OpenGL answers. Then we should be more or less featured between 2.9 and the framework support and then we can start debugging because there are lots of weird little issues that are probably caused by weird code inside Krita. And then we will start stripping the cleaning and refactoring. But we are going to not go to a refactor before the port is completed. So probably getting rid of KDNX4 but resolve must not be given that because KDNX4 is a built-in cooker triggering a good person to KDNX4 especially the Q4 version of Krita already doesn't use KDNX4 anymore. It was not supportable on OSX at least not without making the user do weird stuff. On OSX, user opens from this image, grabs the application folder, clicks it and wants to start painting. And any extra step is a step too much. I wish I could figure out a way to make it as easy to install Krita or KDNX as well. And actually, Norwegian guy is working on that. You can make the portal application easily. These easy installs of Windows are called Oslo applications and they are just a folder which you unpack. There is a version of that for Firefox and also for LibreOffice so you can probably learn something from those guys. We've got that for Windows. But we haven't got it for Linux yet. Of course, that means that I'm maintaining the support of Krita which is something you really don't want to do because that's Qt 4.6 and that means a lot of work arounds. Dimitri maintains the Ubuntu PPA. There's a guy keeping it. Then Lanier maintains the OBS build for OpenSusan. I'd really like just one installer, a dot run file that people download, just like they have with Blender. Because like I said, I try to copy Blender. I copy what Tom does for two reasons because it works for Tom and because my target user group is used to it because they're used to what Blender does. That's also why I love fabricator. I'm a little surprised that we have your own PPA. We have a really good relationship to the DLDAFs. Similar to OBS Susan, I think they should be able to provide that for you. The main issue that we actually started out using the Project Neon stuff to make this PPA. The main issue is the LGS releases. For what I understand, and correctly if I'm wrong, it wouldn't be the first time. Using the regular Wounder Backports PPA, you won't get a backport to the last LTS. And artists are weird people. They like stability a lot and they want the latest features as well. So they want to have the long-term support version of Wounder. That's fine for them. And they want to have the latest freedom. It's actually amazing how many of these artists have built a web browser? Can the owner of this laptop please help me? It's amazing how many artists built Krita from source every morning. They get a built Krita, recipes, tape webcast, and coffee, start doodling. I want a web browser. And I'm just searching for Krita cat guide as in pussy cat. No, not for a guide, but for a cat. Building Krita only looks for cats. And then... I scroll down a bit. This was done by David Rivoy. He first had to set a script to build Krita, and then he decided that doesn't work. I'm going to teach people to use Krita. And he illustrated it. Can I scroll? No, that doesn't work. Just scroll please. Isn't that cute? But there are more illustrations. How to build, how to get your source code from Git, how to get the right Git branch, configure, install, building. There is your beautiful new Krita. Make sure it can run. Be happy. I'm really left handed. And then we developers create awesome new feature. We do that all the time. Our monthly releases are feature packed. So he wants to have the new version. And is happy. But sometimes we break stuff. We do that all the time as well. So he has to figure out which is the last revision that actually worked and go back. And he's happy again. Thank you so much for this great short history of Krita. I hope it keeps going and going, growing and going stronger. Thank you.