 My turn to say a few words. As you might know, I'm Peter Lehhorston, one of the guys that started this schooling project, and I've been the technical coordinator most of the time. I'm going to give a slight update on the status of schooling as we see it today, and then we'll get the few reminding issues fixed this weekend or next few weeks. Next slide, basically. So, School Unix has already told you it's used all over the place. We have development teams quite a few places already. The German and French team already came there in introduction, mentioned the Greece team with Markus Konstantinos in the lead, funded by the Greece government, trying to make an office distribution based on School Unix, taking the network architecture of School Unix, taking out all the educational software and keeping the office tools in School Unix. And seeing the result, yes, but they just started, so I'm quite curious how that will go. We have approximately 890 packages, I think, on the CV, between 80 and 80, perhaps. And most of them are already in Davion. We released the Vidi version with approximately 1,000 packages, and approximately, I think, about 50 of them are missing in Davion. So we managed to get that number down to about 10 packages, and we tried to get it down to zero. We took over the Debian Edius project a few years ago, which was done by Raphael Herzog in France, and he ran out of time, so it was dormant for a few years, I think, and we decided to revitalize it and take over. We released the Vidi version of School Unix in 2004, and we're working to get the south-based version ready for next release. And as I said, we are a Davion project, we are trying to get everything into Davion because that gives us quite a few extra advantages and extra benefits. We won't have as few special packages as possible because every special package increases our maintenance, so that we do not have the manpower, and we do not really want the manpower to maintain a parallel distribution. With the new south-based version, we will move to a new KD. We will make it possible to use a newer kernel. We have all the translations available for Davion, which is, I think, like 30 languages for the installation and for KD, and quite a lot of the user applications. Of course, there are a few applications still missing, translations. There is quite a lot of other news stuff in there, I didn't put up the complete list, so I'll just give a few highlights next slide, please. I want to stress this again. We need to get everything we are working on into Davion, and we need to make sure that the packages we need in Davion stay supported in Davion. Recently, all the Davion packages in Davion were thrown out, which is rather sad because we have been deploying and teaching people schools how to use them in administrator systems, and the Davion packages have been running for two years, and no one is maintaining them properly. And finally, the maintainer decides, enough is enough, and asks the archive maintainers in Davion to produce them. So, from the last week, I think, there are no more Davion packages in Davion. And why do we want the packages in Davion? Because it increases quality, not on its own, of course, but to get packages into Davion, they have to have a certain quality, they have to follow the Davion policy, and if they follow the Davion policy, it's possible to upgrade them. The packages will be integrated with the other packages, they will work out of the box, everything will be very nicely integrated with the whole Davion system. And that increases quality of the packets. We also get a larger visibility of the packet, more testers, more users, a lot of developers looking into the packet if that's built properly with some strange architecture. It's probably a bug in the source and it's not something to be ignored. We also get security fixes from the security team in Davion, so we didn't have to provide secure patches ourselves that uses our workload, and we also get to use the Davion infrastructure to maintain the packet and keep track of bugs and all that stuff. And we also try to push our changes back into what the Davion community normally calls up-screen, which I guess is an image of how hard it is to get stuff in there. But the point is that when we do KDE translations, we give it to the KDE projects, and after a while it shows up in the Davion packages, the Red Hat packages, the Dora packages, the Sousa packages, all the other distributions of course. And our goal is the Davion packages, but putting them into the KDE project instead of putting them into the Davion project to make sure that we increase the visibility of the translations and all the changes we do and we also make sure that the next version of the packet includes them as well, so we don't have to redo the effort to put it back into the packets every time there is a new release. Yeah, I mentioned the Davion issue. There is an early old project to maintain Davion and beyond that, I think his name is an Indian Davion developer who was trying to maintain Davion for a few years, asking for someone to take over for a few years and then finally discovering that the team he had founded early on didn't really consist of anyone but himself and then decided to pull the plug. So there is a project there, the framework to do a team development and maintenance of the packet is there, but it needs some team members to actually be a working team. The French team has several educational factors, like that to get into the distribution, but they are not in Davion yet. We need to work together to get them ready and I suspect there is a lot of hard work left before their policy compliance is ready to go into Davion and I hope we can get that done. The boot system in Davion Edge changed a few months ago to speed up the installation and reduce the amount of boots required to get the install system up and running. They moved all the installation hooks from one subsystem to the other and we used to use the old hooks, which are no longer there. We need to rewrite our stuff to use the Davion installer hooks instead of the base context hooks, because base coating is being removed again. But isn't this a post-sage item? Now we concentrate on Davion's store Linux for such and this will later work. This will take effect when we will go for Edge, but we need to pay attention and if we do modifications to the boot for the installation system, we should try to make it work for Edge too. So we should be aware of this. Next slide. We recently replaced LTSP from a version with ripped out Red Hat's release basically with a few modifications to a version based on Davion packages. This saved approximately 50 megabytes on the CD, giving space for more user applications, but it also gave us a slower boot of the clients and requiring more memory. With the new encrypted connections between the client and the server, the login takes quite a long time as well. We will have to find a way to solve this. There is also sound support being worked on. The new version of LTSP is a cool production of Ubuntu and Davion. We are working together with the people in Ubuntu developing LTSP and they help to fix part of these problems. We just need to pull them back into Davion. Of course, as it's just nothing, but hard work left to do a little rocket science research project. We just have to sit down and spend weeks and weeks to fix it. We lost quite a few user applications from the CD between the Ubuntu and the server version because all the packages increased in size when the compiler changed. That's quite annoying. I think they became like 30% larger all the packages. We need to find a way to select the good packages because at the moment we probably have a few things on the CD that's not being used and is not really fit for use in the school. Of course, we do have quite a few good applications but we could have more of them or we could at least get other applications that we could identify the craft and get rid of it. That's what I am. We'll take a deep tour and demonstrate the new popularity contest. Click on the link. One way to help us find out what applications are actually used in schools is to install the popularity contest packet and configure it to report the packages used to our system. This is Thurman9Machines reporting their application usage and if you have a look at the go-for-the-unknown-vote link this one. This showed the list of non-deviant packages being used the last week in one of the Thurman9Machines reporting to popularity contest. LTSP core i386 is the old LTSP packet but it's too surprising that it has quite a lot of users. AcroRidim, the PDF reader, Java, Paladar Phoenix, yeah. So these packages we should try to get into the end and I've been spending quite a lot of time trying to get free Java versions up to speed to be able to replace the seven Java which is because of the license, number three. Better PDF reader obviously is something people want. We need to look into that. At LTSP we are working on Phoenix and Autolev Foghen and a few others are working on getting into the end and players have the same desk. All the others we should look into and the point is that if you are using a packet which you want to have on the school in Obsidian if it doesn't show up there it will probably never get there. So if you wanted to show up on CD make sure you run proper identity contest and report your user packages and then we will track them and try to figure out how to get them into Debian. Of course we also use the same proper identity contest list for the Debian packages to decide which will go into CD and which will be thrown out. Go back a few minutes back to the slides. Make sure it says the wireless network being very slow down there. So all the issues left to be fixed there is a few of the services in the I don't think you went too far. What is it? That's a problem. It's in the history. Why the hell? I don't know. You can work there, you should know. Anyway, there is some of the services that are not working correctly at the moment. We have the monitoring system the Nagios configuration is not perfect with that way. F11 to be to the screen more. F11 functions the key. Unpitched out, we have some space. Back one. Back one. Yes. NDP I think was fixed yesterday or today. Nagios needs more work to be able to monitor all the services we are installing. Email system, I'm not sure if that's working correctly at the moment. And we are looking into and testing all the other services to make sure they are correctly configured. Also working on the diskless workstations that will not be included in the default installation but we hope to have a script provided so you can convert the normal LTSP client into a diskless workstation. The difference between an LTSP client and a diskless workstation is that the diskless workstation runs everything on the server while a diskless workstation runs everything on the client. But they all use the same file system over the net and behave as a normal maintenance replacement unit. But you get USB support you get you can display videos as long as you have a fast enough client because it doesn't decompress on the server and try to push it back on the network more surface than locally on the client. It's already implemented but we need to do more testing because it's ready to work off the box. Next slide. One problem that keeps coming up is the configuration. And the task of the schooling project is to provide answers to all the questions the users and the administrators don't want to see. They don't want to spend time figuring out which is the best food feature around they just want something that works. So we make the tough decisions for the other for the users. We hide complexity and I know quite a few developers prefer to leave decision making to the administrator of a local machine which is fine with me but it's not the thing we are providing. So we are working to make the packages indebted and pre-configurable in that we can set the configuration at install time and make sure this works across upgrades and hopefully also after if you do install packages after the first installation the way we do this at the moment is using them for preceding which is providing answers to the questions the installation system would ask during installation and make sure the packages think about at installation time and do the right thing. That's the preferred way of doing it but not all the packages we want to configure support is. We also rewrite configuration parts for after the packages are installed this breaks upgrades we provide replacement configuration parts this might break upgrades and we also provide overloaded configuration which works perfectly for some packages not all KDE for example we make some hooks making sure that KDE will look in an extra directory for configuration and we provide a lot of files in that directory to set up email configuration term of the warning that the sound device is missing or printouts and stuff like that. One way to solve it is going for new level configuration but that needs to go through upstream we have to change the behavior of the application itself to get that work so that's the long way the short way might be this KDE toolkit custom distribution toolkit being developed by Senbue in Spain it's a gross hack but it kind of solved the problem by making sure that before the packages are upgraded the old files from the package is reinserted in the old place then the package is upgraded and when the package is upgraded the files are removed and over files are put back in to the old place once so it's hooking into the installation system packages and it seems to be working but it's got to be a better way so there's quite a few things going on and hopefully in the middle of February I'm going to talk to Senbue and a few other developers in the custom level community to see if they can come up with better solutions questions? Do you know that this KDE toolkit I talked to Senbue who doesn't talk very well He told me he was going to do a presentation on it I'm not I don't know the state of the word Any other questions? No? Then I think it should get back to work Peter thank you very much I had to announce another of our program at 14 o'clock we will start a presentation done by this guy there and he will present a German situation in school and free software that's more from a political point of view and so his interest for the PR stuff and this NPR things involved people and different teams are here so now I wish the developers sorry? Mr. Troros will present some technical things and some other things which should be changed in school Linux to use it in schools and I think the developers have to listen This is one opinion I don't know if you want to tell us something I don't want to present it's better you do it by yourself than I and then well I want to have enough time for the developers to fix the bugs this is my main goal so please decide what is important It's a special presentation from the German view of life because in Germany some things are going in other ways and in Norway because we have not so much so far the ministries and therefore because we are 82 million people and Norway is a little smaller and so the administration in Germany it's much bigger and we have in the top level also federal administration we have the ministry for education and we have 16 countries every country has a ministry of administration something goes wrong and how is the real situation in Germany in schools if you want to know something about that and some things you have to do on school because you can change this situation because of the excellent concepts you could listen to my presentation thank you Thank you very much Thank you for listening and I would say the break is now deserved, take a look around Have fun with the work What time is it? 14 o'clock with this lecture 14.30 is the next lecture we have the lecture from Mr. Thorlester Okay