 This is the talk we've all been waiting for, and we have two speakers of twice the fun Budwin from the Krita Hall of Fame, and Sebas, who's been a long time Plasma contributor and has been in many academies telling us all about his exciting adventures and if any of you guys like scuba diving, you should talk to him because he's pretty good at that, aren't he? And hello, who'd like to do it? Hi, so, welcome to this talk, we've called it a racing mobile, it's been a really exciting project for both of us, a bit more building for the past 6-7 months. So, what's up with that? What have we been trying to do? We've been trying to bring Plasma to mobile phones, and we've just seen it. Now, why are we doing this, and why was I so excited that I wanted to work with this? The reason is that the existing mobile phone systems are way too close. It's just like we've heard in Kino today, you've got to listen to it in your pocket, and you can't really do anything with it except what manufacturer allows you to do with it. So, what we've been trying to do is break out of the walled gardens. We've been trying to make sure that we can build a system that respects the crisis. We've been trying to build a system that allows you to inspect and modify for all the stores. And we've succeeded. Alright, can everybody hear me? So, we've been working on mobile topics in the Plasma team since 2009. Maybe some of you remember the Nokia N900, and that was the first device we were running, or we've been building a Plasma UI for. That was a really early thing, so we just tried to get something running on the device. Also, it didn't go very far, and KDs and especially Plasma's ambitions really started to soar when Nokia took over Qt, and it was a period where many things happened. Not a lot of results from that, but it was certainly a good learning period, and it got Qt to the point where it would be very useful on mobile devices. When Nokia, let's say, vanished, we started the Plasma Active project, and its goal was to release a tablet with a novel UI. Well, that unfortunately didn't work out so well for various reasons, but again, we learned a lot about the mobile space. That is to say, outside of Plasma, there's actually really big successes for KDE. First of all, WebKit runs on 90 plus percent of all mobile devices today, and that came from KDE. The Kaligra office suite runs on a shipped product today. That is the Jolla phone, and it's in the hands of users today. Of course, we have various frameworks which are becoming more and more successful, and some of them are also used on mobile devices. So, what's our vision? What did we really want to achieve? Freedom is the most important thing, but it's not just freedom. We also want to reach a point where everyone can use Plasma Mobile on the phone to run their lives, just like we're using Android these days, to run their lives. We want to achieve a product, something that is finished and polished. We want to achieve something that is really, really secure and safe, and has texture products. At the same time, computers are for tinkering, computers are for tweaking. We want to make sure that this, what we are working on, is customizable, not just by us, not just by you, but by everyone who wants to join. Alright, so let's get right to the beef instead of all talks. Let's look at what we have created so far, after that we will get you some more information about how we've been doing this stuff. So, smartphones, this is an LG Nexus 5 running a customized Plasma shell, so it's the exact same stack as we're using on the desktop, but you see it looks quite differently. You can just install this with some parts hardware right today. So in case you want to give me a call, I know not much demo effect there, but this has actually a working phone connection, so you can make and receive phone calls with that. Some of the stuff you see, no in principle, but they have gotten a mobile UI. These are the exact same notifications. So if an app fires a notification through the framework's notification API, it will just work here. We'll get a different presentation because it runs on a different form factor, but no changes in the apps for this kind of functionality. Then we have the Plasma shell, the launcher. It's still a fairly basic thing, but you can already change its look and feel. You can personalize it to a certain degree. As you might have guessed, it supports Plasmoids, of course. Again, here's a different mechanism customized for the form factor. It gives a different user interface for adding a widget. Of course, we had to do a clock. The point really is that the work you've been putting into your Plasma can be transferred almost one-to-one to the new form factor. In some cases, if you're making too many assumptions about mouse and mouseovers and that kind of stuff, you'll have to change some codes. So is this cool or is this not cool? You included K-Runner and that's just a few steps. I really think this is the first time that there's a mobile system that shares so much code with the desktop system. And many of the things here you will actually recognize from the desktop UI. That means two things. They work on a mobile device, so all the APIs and everything is available. They do it just to a certain level, two different input methods. But it also means we haven't changed a lot and really optimized them for the device. We haven't been working for a very long time on this, but we got kind of a really good kickstart because there was already so much functionality that we could just take on the network management widget that took a few years to implement. Well, and we can tap into that work. Of course, a big part of Plasma is the ability to be able to customize it to make it really yours. We just showed some things here. I've seen the video a couple of times already, but I'm still impressed with what it's actually managed to achieve. What we see here is a settings application which reuses KCMs written in QML. So you write a KCM once, and if it's got a responsive layout, you can just install it on the phone and it'll work. And some of the modules we will actually recognize. So again, we've got a kickstart. This is the energy module which is especially interesting on our mobile device. So you get really detailed information. Coco, some of you might know about the application. It's a new image viewer and gallery developed entirely in QML. And again here, only a slight layer changes to make it suitable for a smartphone device, otherwise completely the same code base. You start the same application just on a different device and it adapts itself. Performance is really surprisingly good because this is all hardware-accelerated stuff. This goes through Wayland, OpenGL, and we can make use of a very powerful hardware in the phone. We did also borrow a whole bunch of apps. This is, for example, the web browser which is also used on Ubuntu Touch. It works. It's another one from the Ubuntu Touch project, a weather application. It's one of the things that we feel is really important. This is a system where every application is welcome. So of course, Ubuntu apps run. They are really cute in QML. Yeah, this is not using Mirror. This is using Wayland. We also included some time-wasting applications. This is not my real score. Normally I score in the 900 or 1000 points range. So this is the hardware you get from this model. That's what I mean. So we decided that this project has a Blue Systems internal project because we didn't want to put out vaporware. We wanted to get it to a point where we could say, we can technically make this work and we wanted to get it to a point where it's actually useful for other people to join in. We don't want to get questions. This is really cool. How can I get my stuff running on it? Maybe wait a few months. We're still sorting out some deployment issues. No, you can test your stuff today. If you have an LG Nexus 5, you can just install and run it on it. We also have ExoPC images and x86 packages so you can test this stuff on your laptop and we can help with deployment and testing on the phone. That said, it was an internal project, but last week I've moved all the code that is actually special to the phone onto KDE infrastructure. We put this on KDE infrastructure. We put this under KDE governance. We really want this to be an openly developed and not corporate controlled project. We also don't want this to look like, well, let's hack this up in a garage in between dives. We want this to be a commercial, high quality product. This has to become a viable system that people want to trust and not something that's a really hobby. That is to say, if your hobby is hacking on software, this is an awesome target because it's all open. Also, we feel that we as a free software community especially KDE, we have the mandate to create software that respects and enables your privacy. A lot of I said earlier that you're pretty much caught in this, do I want features and shiny? Yeah, I have to trade in some of my privacy for that. Please don't. We want to offer an alternative that respects your privacy and that allows you to do with your data what you want independent of the features. Also, this thing is built around Qt because we love Qt, because we think Qt is the best solution to build applications for mobile devices and it also really ties in with Qt runs everywhere and everything should be running on this phone. So, the milestones. We've reached milestone one. We have products that people can install and developers can work on. That's already a new step. It works. The video you saw was most faked. That was what you can do with your next slide today. Next step, dogfruity. Actually, you can already use it day-to-day for technical and advanced users. That's something that we will be developing in the coming months. Then, production ready for a really wide audience a truly useful and competitive product. I have to say that this is really a mission goal and a really mission step to it because what you've done is something next year. With that said, there's a lot of work ahead of us to summarize where we stand now. This is not something you put in the hands of your girlfriend if you're interested in a long-term relationship. Don't do that. Don't give it to your mother. You should not be happy with it. You can try it though. You can have fun with it. You can help us to make it good enough for yourselves because we need to make something good enough for ourselves before we can make it good enough for others. It's already really a lot of fun to play with. It stretches what we can do with KDE and it also stretches the scope of KDE. It can really help us to get a new sense of relevance in the KDE project. So what does it do? It runs a kernel. It boots a shell, a plasma shell. This is the same binary that runs the desktop. There's no patching, no code changes. It just loads a different UI suitable for the phone. It has a dialer. It can make and receive phone calls. Some strange people still actually expect that from smartphones. I don't know. We're also using software that we didn't do ourselves with really tapping into other people's resources. We want this to be a project where things come together rather than where we're saying, yeah, but this was not invented here, so we're not doing it. One really good example is the Malit virtual keyboard, which is a very mature thing, shipping on real products in the market today. You can briefly see it in the movie. It actually got a nice plasma skin, so it integrates visually very well, and it's not something we cobble together just in one of Marco's sleepless nights or something like that. We also want to bring in existing functionality, so all the good stuff from the desktop is transferred to the phone. It really gives us a head start in terms of features. You can really console it, so you've got your terminal. What else do you need? Yes. It runs various applications, so we showed some applications in the demo video, but some people do not take mobile phone platforms seriously before they have passed. We have 100,000 applications running on this. Any applications either in the Kubuntu repositories must be more than 100,000. We're good. We can run those. We have support for X applications, so we run Wayland with X Wayland. You can start applications that haven't been ported to Wayland yet through the X Wayland stack. Of course, it's a Kubuntu system. There's Debian underneath. You can run Debian software on it. We also went a little bit further and implemented support for click applications. Those are the packaged applications for Ubuntu Touch, the weather applet. The weather application I was showing earlier is an example for that. That probably gets us in a range of 100,000. One thing we are going to definitely cross this magic boundary with this support for Android applications. We are working on that right now and we expect that within a few weeks, months, depending on when we get some technical things sorted out, you can run Android applications on this. That means that you can give it at some point to your girlfriend and she asks, can I send you what's that message with it? You can just run what's that Android application and it will work. The Android bit is quite exciting because this is going to be a completely open stack. YOLA's selfish folks also support Android apps, but they use closed source components for that and this will be open, everything will be open. Yes, so you will be able to run Android applications not only on the Plasma phone, but also on the Plasma desktop and that's one of the really, really important things for us because we're not saying the desktop is done now, we're doing something funny new, but we're adding the mobile platform and UI to the Plasma offerings. We are continuing to improve the desktop. We also want to make a dent in the mobile space and conquer these devices. So technically the interest you get is we started out building a Plasma phone, in the top of Mer. Mer is basically the old source base for service from it's market-based system, it uses Wayland and we were using Queen Island and keep it in Plasma. Then we decided because this is a new system project we are going to move to Kumuntu as a base and use Kumuntu phone. That proved to be really challenging, it was really challenging to get a cute label to up and running the top of the boom from where we access the stack because all these three open and open phone systems they are all based in Android and somehow you mix a real-life and an Android together in one way or another using the language called the Pirates. So we were really struggling to get this to work to get even a single image rendering on screen until Martin Hasley said, well, how hard can it be to actually use Queen in here? And that's when that moment the project really started to fly. Ported Queen to the phone, started writing a back-end after talking in Pirates and we were actually getting graphics. So one of the most exciting things for me and all these systemers is that we now have a phone system that runs quickly and where it will talk to Kumuntu from. All right, so let's see. What are we going to do and why can we make a difference? Because some people will say yes, there's mobile phones and those are free enough. In our opinion, they are not free enough. You have to trade in your privacy and even worse, sell your soul to mobile vendors. That's the reason for us and that's the reason for hardware vendors. And let me tell you, they are not happy with being under the thumb of Google and getting source code thrown over the wall every once in a while. We have an active development community who produces software for end users. We have an awesome infrastructure with our participants deeply caring to make us productive. We have a governance model that is actually very suitable for creating software and we have proven that in 20 years of being a successful and reliable maker of mainly desktop software. We have open and proven development processes. We know how we work together effectively. We know how to get things fixed. We know how we can create processes that improve the quality and that get our software out to the users. We don't have to reinvent the wheel in all these supporting factors so we can really concentrate on making good software and doing so even in a fairly short amount of time. And most importantly probably we have a proven track record of being open and respecting freedom if we put out something like that we haven't benefited an advantage in trust already. If we talk about privacy we can do so with a straight face. If Google talks about privacy not so much. So with that said Blue Systems is working on this but we don't want it to be a corporate control project. Also we want to involve the wider community because it's fun and it's going to be rewarding and it's also a very important project to work on. Both for Katie as a community but also for the users who are right now not really having options if they need the privacy. And it's not just end users at home. If we want journalists to be able to communicate in a secure manner to discover things and not have them be discovered by governance and in this chasing game if we want to enable the Snowden and Apple bombs of this world we need this kind of platform. We need you to help us create this kind of platform. And with that I think we are coming to the questions part of this presentation. In our best Spanish. How do you make sure that an application can just track you everywhere with GPS for instance and things like that? A lot of that depends on the application but we also have to admit we are in a very, very early phase of this project so we have no good answer to that. Like for many other technical issues as we said we have a lot of technical issues as we said this is not ready for end users yet and sandboxing of applications is one of the topics we need to work on. It will also depend a lot on the application on the type. I'll wait for that. Okay. Okay. So let me write sandboxes on the application because that's built into the system. What we've got here is Linux. It is Linux. So you could try to use as many as you can to do the sandboxing or you could try to develop a new model but it won't be exactly like this in this basic unit system. Do you know where I showed Plasma Attachable while at the end you showed Plasma Dot Multilabor? Which unit? The edge. Which one is the dot? Is the... Is working on... working to other common devices except for the XS5? No. But a very good and important point. Two things we haven't mentioned. What we are presenting here is a reference platform based on Kubuntu. This is not a Kubuntu-only project. We want other distributions to create a Plasma mobile platform as well. If you want to get this to run on a phone running Fedora, we are going to help you. We don't want to create one single distribution. I think we want to create a platform that you can run on your mobile devices on the US. On the US you want it to run on and we will help you to get that running. I have two questions. I am from Taiwan and I remember a few years ago maybe 2011 or 2012 there was a demo of the KDTel in the Coscom. Coscom is the biggest of those companies in Taiwan. There was a demo of the KDTel that actually was interesting. That year, KDTel could know both demo and tel. But KDTel is more complete. But in Taiwan we have I don't know if it is a production but in Taiwan we have no any further problems about that. I don't know if it should be succeed or not. If not, how do we see your work Plasma will now succeed? That's my first question. And the second question is similar. Compared to now we have besides enjoying the iOS with a 5.0 OS who want to touch what do you see your advantage on the 5.0 OS even enjoy the iOS 2? Alright. So first of all this has actually taken a lot of the code which was also making up the Vivaldi tablet. This code has been ported to the framework slide. This is not running X11 anymore this is running Wayland so the software stack is much, much leaner and much, much better. Then another important difference to the Plasma Active and Vivaldi projects is that we have decent hardware it runs on today. The Plasma Active project was hampered by we couldn't get it to run on a real ARM device so you usually had to log in but the phone said I carried a lot less luggage to this academy than to earlier academies. That makes a big difference. And what is why don't we just say this we want to touch already and we'll find that well I think the past weeks have shown pretty well that we want to touch is not exactly a free and unencumbered platform it's corporate controlled and there are many questions around trade which may end up making it not a much better option than Android for many hardware vendors. So we do have something to bring to the table. Also if you compare how long it took Canonical to get something running on the phone and what kind of resources they put in and how long it took us but even to get even a prototype running let's interpolate from this and let's meet again at academy 2016 and see why we stand. We have time for one more question. Sure. I've got two questions. The first one is hopefully very simple. So you said that you're still depending on the purpose. So I guess the drivers are still preparatory. Can this change a little bit as if it would just can it? Yes, absolutely. So that's this diagram which let me not pull it up right now. It's to save some time for So Quinn uses this device to get the graphics on screen. But it's Quinn. It has a backend for DRM so it can use any other drivers and send code based actually also runs on my tablet. So we've hidden this difference neatly away. Yes, we don't have a free graphics driver for our reference device. Yes, we want to. Good. That's excellent. The second question is this reminds me very much of the Gemini project and some feature that I've seen in Plasmoid since very, very early 4.0 which was that you can share Plasmoids across that work. So is there any is there going to be in one of those systems which everybody is trying to do where you have it but as soon as you come on your large screen and a keyboard it suddenly turns or a projector it suddenly turns into a proper desktop or will it be able to read the mind of my laptop and show me the relevant Plasmoids for example in the calendar? It may be able to read your mind that probably won't send this data to random services but yes the use cases you are talking about we haven't worked on them so far we've been concentrating on getting the basics to work we definitely want to turn this into a device that depending on the use case you run it on but probably for the Plasmoid case much more important is you usually start something on a device and the device doesn't change so it's not critical right now to have it morph into something different but if you plug something in with the Plasmoid architecture this has been planned from day one so yes it's fairly easy to support these kind of use cases Thank you Are you already going to find the states? That's all the time you have They can be around in the academy so you can find them