 about Malix. I did four semesters Chinese, but I'm not very good at Chinese, but I was told that may mean something like beautiful in Chinese. So I have no insights into Chinese, but maybe can give you some insights into Malix. Oh, this is my name, this is my email, Twitter. So where to start? Maybe start here. Looks familiar. So you have like this typical Ubuntu or Deviant style boot up screen that we customized for Malix. You can install the system on your hard drive. You can run a live CD. But that's actually not the approach we want to take. So what's special about Malix is not that it's just another Linux distribution or just another customized Linux distribution, but it basically starts with a use case. It's a desktop application, desktop distribution. It's very special and for a very special purpose. And the purpose is, in short, think of a conference organizer like FOSS Asia and want to set up in a conference space 100 internet terminals. And there should always be the same software on each device. We have certain requirements for public computer where you could access the internet. So in the center stands the use case in this case. And we reuse all existing technology. We try not to invent too much on our own. We choose a very simple desktop environment that is LXQT We reused the original build scripts from Ubuntu before Ubuntu was existing as a starting point and try to really work on just these small and feasible configuration items that are relevant for our use case and always try to turn development onto the use case to satisfy the user needs for this special desktop purpose. And the idea of the development is that we apply continuous integration. So we try to, every time there's a change, we build the system again. And we really have this kind of approach of an agile and use case driven development process. So there's actually not only Mailex, but the important other tool is the Mailex configurator. Customizer, this is a very minimal viable product. It's a web app running a container and basically ask you for your email. You can say what's the name of the event to pop up when you open a browser on this customized machine. You can upload a wallpaper that's like the wallpaper of your conference. These are just example settings for the proof of concept. Of course we want all the other features necessary for a configurator. What happens then? You press the build button and then you get an email. I thought it was later, but Tarun made was 10 minutes later, you get an email. And this is the email with the ISO build. So you download an ISO image and then you can install it on the different machines. And the idea here is to keep it very simple in a way that the person who actually creates this customized distribution doesn't have to care for the technical details, doesn't have to run any build scripts, doesn't have to know much about technology, just has this web interface and everything else works in the background. The mail is as such where you run the build script for distribution yourself and this is almost similar to say what your distributors do when they compile a new release of their distribution of the software. So what we found important in the process and when we evolved all the different people is to follow the FOSS Asia guidelines for software development using existing infrastructure, using infrastructure that is easy to maintain so we don't have much administrative overload. Reusing existing tools like using JIT up for communications is just simple because we can use it as an external service, we use JIT up for the bug trackers and so on so we don't have to maintain any infrastructure. And yeah, to try to at least this was the stage where we always build and when we start to correct it. The other point was when we had our students working on the project and they solve tiny tasks always encourage them to document how they solved it, what they did and writing blog posts. We found it in the GWSOM of code process very important to encourage people when they contribute code, when they send a poll request, also write blog posts about what they did, how they solved it, what alternatives they considered because this is very fruitful and we discovered that the students that were very good at writing code were also the ones that were very good at writing blog posts surprisingly because the stereotype at least in Europe or in Germany is either you're a good coder or you're good at documentation but in the first Asia sphere it's a bit different so and the third thing as an observation from the Mellings project which is a bit like a training ground also for our developers from India mostly is that we found it very important with a larger team to just walk through the code much legacy code from 2012 and look what is it actually about and walk through the code together jointly and using teleconferences and just trying to improve our common understanding of what's going on and what may be obsolete and what's actually applicable so here are some links there's like three repositories one is for the Mellings the Mellings as such for building it these are simple shell scripts that pull together the packages they open up a CH route where some other software gets installed and some configuration happens and then we have in the Mellings the generator branch that's important for for generation of this taking this this image for the for the Mellings generator then we have the web app that's also exported the Mellings generator as a separate repository and the third one that was worked not last year but the the year even before is the Mellings system log because when you have a desktop that you provide to a general audience what you want is you want to lock the system down so that all the casual users that are coming with the different requirements the different interests the different languages and so on cannot destroy the computer or play around the way so and main discussion takes place on a jitter channel in in the first age just a leaks so how will we continue with the Mellings project Tarun will go more into details one of our developers but soon but I think for four projects I would say could be singled out and are relevant for the upcoming Google summer of code season and one is the roadblock project like you know it's more or less experimental code it's not very beautiful code in any way or maybe partly because it's a configuration task not too much code at all but one of the roadblock projects is of course we have used these old scripts from 2012 and quite useful to put this into productive use to yeah migrate everything to 64 bit there's no standard we have to do some some modifications there so someone has to work on that and there are also some some other blocking projects like how to kill desktop notifications and other details the second project is something something new we want to do based on Mellix but it doesn't matter whether it's the target platform is Mellix or just an ordinary Lubuntu with with XQT but Mellix is like the reference platform for it and this is Susie desktop integration you had all these talks about Susie and Michael Christen can explain you the details on the conversational server and all the different clients that are right now available but what we don't have yet is something in the Linux space comparable to Microsoft Cortana for Windows like desktop integration of conversational features and this also cannot totally protest on on the server because all the different all the clients are different so we have some kind of API to steer some desktop specific options by with a conversational interface and provide some interface to get feature parity with Siri with Alexa and so on so all the hard work all the stuff in the background was the Susie AI server is basically done but now we have to put it more into use and support the client side features and this is what this project is about and we hope to get a sum of code student who wants to work specifically on desktop integration so the and the third project of course is to improve and expand the Mellix generator now that the basic say if you if you know how to configure the wallpaper then you can also configure everything else this then becomes pretty trivial so we want to expand the features and make this new more useful and there's an idea because right now the build process is pretty pretty ugly using a bit of a heckish shell script to use a next generation built build process and this is prototype by harsh in his fork of the Mellix main repository so I just had a look at it I don't I don't know I haven't tested it and I don't know how much work it would be to actually make it work and the seemingly replacement but this looks like the way to go from a technical side so maybe you so yeah good why desktop when everyone's moving to the phone that's actually good point I would say that the desktop still has a place the desktop is still important I mostly use the desktop for all interactions and I am aware of the mobile hype I'm aware of the mobile integration but still say the Linux desktop is something that never happened and the closest approximation to a system that works out of the box is actually from my perspective to go simple and we did it a few years ago when we started with this LXDE project and try to try to get a more simple go into complexities that just provides the basic functionality and you say okay this is like a drop in replacement because right now public administrations they are not going to tablets they're still using desktop computers and they will continue to do so for all their services but say public administrations they were considering to switch to Linux and it was a time where I could say yeah let's try it let's do it let's migrate to Linux but right now you cannot say this anymore at least I wouldn't say I wouldn't recommend a public administration to switch to Linux on the desktop simply because these desktop environments are too complex not stable enough and we had some say my migration disaster so either your software gets completely outdated or yeah useless or yeah unstable and dysfunctional so I'm using the desktop Linux desktop for for ages almost 20 years now and I'm very satisfied with it but I don't know if I would recommend it to someone else so what's really important is to keep it very focused and in a way say for me the perfect for me the perfect desktop is like like the perfect servant a perfect servant is a servant that you don't even notice you just see oh there's a glass of water exactly in the moment when I want this glass of water and when you talk too much about the desktop and are too concerned about all the features and so on and note you've noticed the desktop then something something goes wrong but I I guess say for for general user who's also experienced in this Microsoft sphere I think Lex QT is quite good approximation as a desktop so this is the software I would right now recommend or all to go with cinnamon of course because the reason why I ask you that is I actually have and almost everything had to be installed by command line there was not a single like if I wanted to say go to a Taiwan office right I guess this is a general the general question that applies to all Linux distributions in a way some some some are very friendly some work out of the box sometimes don't work out of the box personally I think if I can get some command line solution it's it's better for me I don't want any graphical installers or so simply because when you think of a support context when you support a user and tell him how to install a driver and you you just give him give him the instructions for the command line it's something like okay use this shortcut fire up a terminal and enter these three commands and what will be the output or you type in this command what what does it show to you this is good for good for interaction for for customer support but if you have graphical interfaces I mean try to explain your mom on the telephone and then you have to click the button there in the menu so-and-so it's it's really it's it's horror to to do to do remote support on graphical user interfaces and solving all the issues so personally I prefer very much the idea of frontline interfaces for configurations if there has to be something configured and yeah the other and and regarding say of of hardware I mean this is not genuine to the platform say if you buy a computer in a shop with Windows pre-installed then it's in the responsibility of the hardware manufacturer to make it run on that machine and it probably works because it's shipped with it okay on the other hand I problems when you run older hardware when you run hardware that's not certified for your machine or that was certified for the previous reason for the previous machine especially when you try to use older printers with newer machines at times the printer manufacturers they don't want to support the new operating system and you run into troubles and just just an example of from my practice say I wrote I got Bluetooth headphones so you cannot operate them on the Windows 10 you cannot operate them with Mac OS and with Linux it was very heckish and complicated to get it get it work but now it works and then there was a kernel update and now it works seemingly as out of the box so so I guess that the Linux approach still is the best because you have such a variety and such a richness of hardware support the point with Malix is that our users shouldn't have to do anything with system configuration so if if this system doesn't work on a on a box that we put somewhere then there's a problem with this box so then we cannot proceed with this but this is just just a use case for for development I wouldn't recommend right now to use Malix for productive reasons in any way but thinking about the use case working on the use case means thinking about in this specific context what do users need like when you have this for a conference space then you come up with different requirements that you usually don't have like support of multiple languages if you have an international conference some speakers will be Japanese some will be Chinese so we have to have Japanese and Chinese out of the box and the next person that comes is a Spanish citizen so these are very special requirements or maybe you want to block certain websites and her block list so though they don't use your conference space to access certain websites that may be inappropriate in a in the context you wouldn't have wouldn't want to have this at home of course but say in the conference space I understand that there should be some blocking of some services and so on so very deriving from the use case what do we need in order not to confuse users in order not to create too much overhead technically Malix is right now I know that there's some some in the projects who are very much into our Linux but it's basically built on a on a Debian or Ubuntu infrastructure so we don't even we don't even take our own kernel so we even take the Ubuntu kernel which technically doesn't even make it a distribution because if you don't compile your own kernel then it's no distribution as you know but that that's not the real case the real case is the use case and to trying to develop all the configuration so everything get things get set up out of the box and the special needs of a user are satisfied here and yeah why not why not a mobile phone actually say in these conference spaces where you have like this internet terminals or in a hotel this was the other use case Malix was previously called hotel loss so the idea was yet to create a distribution for for for hostels and and for these purposes so but but this is basically the same a device the desktop that accessed by multiple persons temporarily persons who don't own the computer persons who just use it for its and want to leave no traces there you have to provide some security and and so on and settings shouldn't confuse users in any way so no pop-ups in in in Chinese and no bullshit okay yeah maybe maybe Tarun can get get into the details by your words yeah yeah the software is right now mostly in experimental stage what we have say dedicated packages of course but they some of them don't adhere to the the deviant standards yet as someone pointed out so we also have to work on that no no we don't want that but we had the discussion yeah but but actually I guess right now we have no no users other than the developers because yeah we were still working on this project for also as a kind of educational project or for trying out some some stuff how to build build from from an app using automated services building ISO images and that's that's kind of the fun okay so I have to