 So good morning, everybody. This is Andrew Steele with his talk, While Running a Blend. I hope you enjoyed the talk. You have 45 minutes. OK, thank you. So thank you. The first reason I told you, officially, just to travel around the world and get invited for this. And but there are other reasons for running a blend I would like to show you. And first, I want to give a very short introduction to the day we are made because the day we are made in principle should not be the topic of this talk, but it should just serve as an example. Then I want to give you some introduction into the blends concept and why I consider the day we are made as a model. And then I will show you some graphs which somehow proves my thesis that a blend can be a successful approach to cover certain user interests. But now for the way we are made for the history. This image was made on DeppCons 1 in Bordeaux. Two people are also at this DeppCons. This is James Bromberg, and this is Colin Watson. And the other people are not all also active, but several people are not active anymore. That's not the point. Though it was a social dinner, we had some Bordeaux wine, some inspiration. And the day before this dinner, I have observed the French guy translating an installation instruction for a medical program. I said, well, translating an installation instruction, what's the sense of doing this? We should package a program that nobody needs to care for the installation instruction. The user instructions, this makes sense. But installation instruction, no. And we discussed this, inspired by some wine. And then we started to create a talk. As usual, you have some wine. You are on Debian conference. You create a talk. And this is basically the first talk I had about Debian Mids, and the name was not invented. It was nine. But you see, it's a little bit crazy idea to work in a field which is not really addicted to free software. Medicine is involved with money. And it doesn't cost money. It's not good. And so it's a crazy idea to try to cover a field which is not really thinking about free software. But it somehow worked. Now, when I had some talks about Debian Mids, or when people hear the sound of Debian Mids, I think it's a practice management system. It sounds logical. But it's wrong. It just contains a practice management system amongst a lot of other systems. And basically, the practice management is only one very small part, because there is not so much you can manage practice. You have a lot of medical imaging software. And we have a lot of this in Debian. And you have even larger amount of software for gene sequencing, preclinical research, and so on. And this is actually the strongest part. And so because I'm working by chance at a medical institute and this biologic stuff is quite used there, not a free one. You're using something else. But I hope I will switch them. I thought it would be a good idea to start with something where free software exists. And we had even some packages inside Debian that exists. And so it is a whole system which is targeting in different fields of medicine. So Debian-Mid is a so-called pure blend for medical care and microbiology research. This is somehow the definition. And the goal of Debian-Mid is to provide a sustainable way to distribute medical floss to the user. And the other thing is we want to make Debian the choice for people working medical care and microbiology research. So if somebody is asking, well, I'm working medicine, what distribution should I use? People know Ubuntu is easy to use because they always are told that it's the case. And it should be normal that the answer is use Debian. That's fine. I can also use Ubuntu because they derive from us and get everything we do. But, well, that's the plan. And the goal of a blend is, well, it's easy. Just substitute Debian-Mid by blend. And it's actually what I've done a lot of times in Debian because we have teams which formerly did not really consider themselves blends. But they have some tools made for Debian Edu, for instance. They made a mid-package approach. It was invented by Debian Edu people. And they said, well, that's cool. What I've done is substituting Debian Edu by blends where it works for everybody. And in the same way, Debian should be the distribution of choice for people working in blendscorp. And you are really free to insert blendscorp, whatever you want to do, be it games, be it multimedia, be it giz, be it some science, whatever you want to do. The framework is there. You can do it. You can use it. And I will show you how to do it. This is my idea of Debian. You have basically people working on packages. I'm putting in a common pool. The user is using it on his computer. And the good thing is that you have people who are responsible. Frequently, there are more than one face behind one package. This is good, which are teams. This slide is a little bit old. This is just my, when I started with Debian, this word net is not addicted to medicine. So this was just my first package when I entered Debian in 1998 when I came to Debian developer. I'm just 15 years. But the user view of Debian as well, this is a lot of packages. We are the packages I'm seeking for. I have very special needs, but we are my packages. And so blendscorp can help. They are focusing the user's interest into the packages. He's really interested. And we have more of them. We have ScholarLinux, Debian, giz, and we have others. So these blends give some focus on the packages you are interested in. And that's the main point. And to refer to Joe's talk yesterday about Debian cosmology, my vision is not so good to see the development of all these branches into different blends could cover the whole universe of applications. If you just think about it, how we can find users. Because for instance, between Debian Made and Debian Chem, there's some connection. We are providing the same packages sometimes. We are not teams fighting for the packages. But we say, well, this package of maybe another team is interesting for a medical user. Or this team, which is more focused on biology, could be interesting for chemical users. Or for instance, the connection could be if there are some chemical applications which can use in schools and so on. So all blends might have some connection points. And if you are, for instance, a teacher and you are thinking, well, my school is using Linux, even edu, whatever. Linux is a cool system. I will install this system on my brother's box, on my father's box, and wherever. So if we dive in different fields of work, people will use it at home as well, maybe, hopefully. And so my vision is that we could cover all the universe with Debian if we try to reach out into certain topics. What's the name blend? People might wonder about this. Well, at first is what's called Debian internal projects at Debian Conf 3 in Oslo. And I think it's not a bad name, but the problem was we have technical projects. And we have user-oriented projects. And we wanted to have some distinction between it. And so we decided to rename the user-oriented things to something else. And this something else was called custom Debian distributions. But this name was a complete failure. Because if you hear the sound, custom Debian distribution, you immediately know what it is. It is taking Debian doing something with this. It's something else in Debian. Because it sounds like something else. And I had talks, for instance, about Debian science. I had all the talk. I had a large slide saying in boldface, Debian science is pure Debian. There's nothing else, nothing added. And the first question afterwards was, why are you doing another Debian derivative for science? So finally, I decided to change the name, because it doesn't make sense. You are fighting against windmills like Don Quixote. So my idea was Debian internal solutions. But people popped up. Well, I have a certain opinion on the name. I have never heard about these people. They didn't provide it any code. They were not showing up in the mailing list before, but they had an opinion on the name. Fine. So we ended up with some votes. Jonas invented the name Blends. And he won. Fine, I'm perfectly fine with this, because Jonas is always right, I know. Now I haven't done recording. You haven't done recording, sure. So the Debian pure Blends is pure Debian. And if you are blending and mixing a little bit for whatever reason. We just settled on with the name Blends. And the advantage of the name Blends is you don't not really know in the first place what Blends might be. And you start reading the documentation. And this is what you want you to do. There are similar structures in other distributions. For instance, Fedora has so-called special interest groups, six. There is a Fedora medical sick, and there is a Fedora science sick. Well, I'm a person, I'm not fighting about against other distributions. And as I said in the beginning, we want to make people that they naturally answer Debian as the right distribution for you. I do it with a fair competition. I work together with them. I'm on their mailing list. I give them hints if you did some packaging where the patches are. And so there is some cooperation. But I must admit that from the amount of packages, from the team behind this. And so we are way more advanced than Fedora in this field. And there is also open through the medical. This is a little bit of a sad thing because I'm fine with their copy stuff. And they copied our task page. I'll show you later these pages, which are listing all the stuff we have. They copied it and put it on a web page. What's nothing behind? Nothing. So if people come to this page, look at it, learn. There's nothing. Then they come again to our page and say, well, I've seen this. There's nothing. And they do not realize that there is anything. So that's a problem. I contacted them, but no response. It is even outdated for two years or so. Yeah. No, I can just say that they can put you in contact with a guy taking care of the communities in the open source at the moment in case you want. Yeah, I have one German guy, a Spanish guy. OK, perhaps this is somebody else. I'm interested for sure. Yeah, so I mean, I know him for a long time. Would be nice, yeah. So I'm all for cooperation and so, but please make some sense with your cooperation, right? And then there's some Ubuntu-made project. I tried to contact them, but I never got any response. And they seem to have some different focus. And largely behind, I was never becoming really clear what they are doing. If I would seriously run a project, Ubuntu-made, I would contact them. I'm waiting for this contact. If anybody knows something, we are open. We are like to work together with you. And there is something similar in FreeBSD. They have ports, and they have a lot of biology packages. Some of them we don't have, but the whole amount is below. But this is also an interesting project for biologists. So Debian-made has by far the largest team and the largest amount of packages. So what's the relation to derivatives? Because there is for sure a relation between blending inside Debian and derivatives, which have a certain focus. And my opinion is, and I always say it in specific Debian-made talks, what we want to provide is maybe the optimal solution and the least minimal work if you want to derive from Debian. So if you want to apply Debian-made in a hospital, a medical hospital, we can't do this. We are just a small crowd of developers with basically no knowledge. But what we can provide is if you are a service company providing services for hospitals, then you can take Debian-made and do a lot of stuff automatically, which is done for you. And we want to do the preparation. And in this sense, we had some very good connection to so-called BioLinux. They are even using our VCS. They are committing their changes to our VCS, and they are profiting from our changes. And we even join ourselves in sprints. We are doing come-in sprints. They are developer from BioLinux, coming to us, working together with us. And they finally profit from our work via the Ubuntu data. All our packages are in Ubuntu, and they do it via Ubuntu. So this is fine. So we have good options, but they are unfortunately widely ignored by derivatives. For instance, there is this OS-Geo, which is in Ubuntu derivative, which tries to get the full tool chains of open source Geo special software. And they do not proactively interact with Debian data. For instance, they don't know that there is a Debian Java team. But they should, because they always had the question, how to package this and this and this. And I'm now reading all their mailing lists, and I say, well, there is a Debian Java team. Just ask your question there. They will help you. I hope this works this way. And it's not only the derivatives fault. Sometimes we have Debian Junior, which was actually the first blend, but it's not maintained anymore. And perhaps we can do something. So there are some problems with the blends in Debian, because it's not so many menpower behind it. But actually, regarding the menpower, I will show you something which is probably interesting. Well, the purpose of a blend is making certain topics hot inside Debian. And we want to attract users working in special fields, because if we get users, we get also developers. And it's also teaching users and developers how to work together with Debian. For instance, in the first five years, I did all the packaging myself. And my work shifted drastically from maintaining or measuring people, sponsoring packages, doing some organization management, mail writing. So it moved somehow. I'm doing totally different things. And my vision is that Debian is just a way to advertise Debian, a blend is a way to advertise Debian in specific work fields. Why do I consider Debian made as a model? Well, it was one of the first ones. And as I said in the beginning, medicine is not so hardly connected with free software. But if even this kind of a leaf topic was able to survive inside Debian, how much more chances should have multimedia games? And so if it would be maintained in this certain way, I want to catch you to do some work. So please don't leave this door without the intent to do some work on this. And we also show how to form a specific team around it, because my idea is you will not get some supporters if you can't show anything. In Debian, you can lead by technical competence. At first, you show some code. And we have some code developer which works for everybody. So providing these tools makes it way more easy for you to help others. For instance, we have now Google Summer of Code projects. That's my student. He will rewrite the stuff we create the meta package, because it had some flaws. And the tools become better and better. They are ready for using. And so we finally was able to push Debian into a field of dominant priority software. And you can follow this path to turn Debian into the distribution of choice for your blend. So how about the tools? We have some tool to create meta package. The blend has basically meta packages. So for instance, if you want to have all the medical imaging stuff which is contained in Debian, you say, up, get, install, and make imaging, and you are done. Your machine is ready prepared with everything we have. And as I said, this is a generalized idea from Debian Edu. It's called Blendstaff, and it's currently rewritten to be even better. Then we have this so-called web sentinel, because, well, my colleagues sometimes ask me, they are not using it. But they should use it because we have everything and we have it even in a better quality than on their other machines with other systems, right? And, well, I consider how to say what I'm doing or explaining my mother what I'm doing. And I invented the so-called task pages of the web sentinel. I'll show you. It is just a list of everything we have. It is, well, this is medical imaging. We have screenshots. You can enlarge the screenshot. And we have information if it's up to date. This one is not. All the yellow buttons means work. So we can look what work needs to be done. It's something to do for the DebCamp or DebConf. And this is just a list. And the list, you see, my intention I left the German translation that you can see. It is translated. We have a lot of this. So it is translated in your language. And if it's not, you can maybe we find something. You can fix the translation. This means fix translation. You can translate. But you can put on the button and you can translate. And my idea is if I was able to attract a user on this page, he is a competent translator. The thing is I started with translating descriptions of biology packages. I'm not a biologist. I'm a physicist. I don't understand the world. I can't translate. You need to understand something to translate. And so the idea is we want to bring together developers and users. And they should work with this page. And some of the work can be perfectly done by users. It is translation and upload the screenshot. And you see if there is a screenshot, you have a screenshot, otherwise you have a button. You can do some work. And I always want people to do this work. So what else? You can see popcorn results. You can see. If there are no DevTags, you can go tagging. Because DevTags is quite related to Blends because it's also categorizing packages. This is what we are also doing. And I also seeking a way to include DevTags more strictly into Blends framework. There are some differences because if you go below, further below, do you see we have a lot of medical imaging packages? We have this yellow section. You can't see the difference between, well, here. This is green. This is inside Debian. The yellow ones, it's not yet in Debian. This might be in official packages, for whatever. And we are working on this in Git or SVN. So if somebody is seeking for a package, and it's not yet in Debian, he can look up this page and see, oh, maybe there's not so much work left to do when I can help. Or you can install something for testing. And so you can't cover this by DevTags because DevTags only covers existing packages. So you cannot always do this. And from the meta package approach, it's done this way. The common database are so-called task files. And task files are used to build meta packages and these task pages. And if you build the meta package, the process looks if this package exists in Debian, if not, it decreases the dependency to suggest. So if there is some in-official software repository, you can use upget install minus, minus install suggest. And you get it as well. So this is not very widely used, but it could work. So this is about working together with people, trying to attract people to common things and show where the work is waiting for you. We have also the so-called bugs pages. This is a collection where I try to get some idea where is a lot of work needed. It's not in these sections because this is not marked. Medical data has something with bugs and the other packages have some bugs. Same here. Take for biology and you see a page where it lists all the bugs affecting biology packages. The idea behind this, if I'm a biologist and want to do some support for Debian and looking for bugs, well, it's most interesting to find them in this category and not in multimedia, where I'm not a competent. So this is the idea of bugs pages. Then also we have the so-called termometron. I'm not really using it, but the Debian GIST people are really, really, we need our termometron. So I implemented the termometron. It is a kind of marking when a package has the same version in stable testing and unstable, then it's green. And if the version in stable is lower, then it's yellow. And I pimped it up with a bit more data because they had it implemented with their tools, but the blend tools are much more powerful because working on the UDD and UDD has a lot of information. For instance, about work in VCS, which was not contained before, which has upstream information. And so this package term was a bit pimped up. And my strategy is to reach out for the teams, which are potential blends and provide them the tools they need based on our framework to attract them and follow this path. So we also invented the handling of scientific citations. I need to switch back to biology because we have a lot of them there just to show what it is. Here you have biology. So here you can see police site. So I'm on a police site. So most of these packages are done by scientists and they really want to get cited. And this scientist said, well, we get another hit on Google via Debian. Well, that's interesting. But what could we do for bringing our package into Debian? And so we can convince upstream developers to build some Debian packaging. And we're mentoring a bit and showing how it works and fixing bugs. And this saves a lot of work. So the same way, trying to connect to the users and connect them to Debian. So we did some subsectioning of DDPO pages. I'm going to show this. It's not that much interesting, but it's new. And we make some use of UDD because we need it for our web sentinels. So I entered some importers for the citations stuff and for the VCS packages. And I also have written imports for the packages in new because we also list the packages which are just uploaded to new and not yet included in Debian. So there was some work done for UDD. We have written a Debian made policy, which is just a document. So if you want to click on this later, then everything what is underlined is a link. And you can use it. This PDF is online. So the Debian made policy is a document which explains to newcomers what the rules of the teams are. And in the beginning, we copied this from the Debian Perl team, thanks to Gregor and others. And other brands like Debian Science copied from us. But this Debian made policy is very good because if there's a newcomer, what should I do? And please read this, how to create an Alliot account. And so something is there. And you can basically copy it and change it for yours. And then there are some other things I invented recently, or mentoring of months was last year. Mentoring of months is, well, how to start. If I'm a free software developer, I can spend my time at will. And my boss will not say me anything. And so if I as some kind of project leader want somebody to do something, I have no handle. I said, well, I will trade my spare time for one month. And you trade your work for one month. And I will teach you for one month. And you can ask me any question, but on the mailing list, and open to learn the rules inside Debian Made. Please ask everything you need. And I will teach you, but you should report what you have done. So this report is something I have some. It's not a rope, but a small thing, small connection that I can make sure this person is really working because there are people showing, oh, I want to help you. And that's all. And so I try to get a little bit more stress on this. And mentoring of months, what does one package actually not yet uploaded? But it's quite a large precondition for a hospital management system called MUMs, MUMs database. We have no competence in the team for MUMs. But this person was MUMs developer. And we teach him packaging. And now we get something done. And this is connecting to users and working together with them. And I also recently started sponsoring of Blends. What I mean is I have learned that the Blends concept, even if I'm talking 10 years about this, is not widely known at but should. But we have a lot of people who are trying to get their packages sponsored. I said, well, if this is in the field of any Blend, is in the VDS of the Blend, is on the task page listed to make sure that the user has, or the developer of this package has understand the Blends concept, then I will sponsor a package. So this was not yet used that much, but perhaps because finding a sponsor is sometimes hard, but I want to make it easy for the Blends. This is the example for mentoring of MUMs. This is funny. This is Luis Ivernes. He's also a physicist like me. But from Columbia, and so with a quite heated tempo, and he said, we are looking for one of you to help us change as a history of health care. So we are currently changing the history of health care. I told him, well, you can't change the history. You can only change the future. But, well, OK. And it was exactly the day when I have written this mentoring of MUMs stuff. And I sent it one day later with a connection with a 10-year birthday of David Amid. And so he got together with this guy. He's really cool, and we get something done. It's really fun. So another thing is, why is David Amid good for David Amid? I did the questionnaire in the wiki. There's a link. And David Amid has currently 23 developers and David Amid maintainers. They are not all that active. So 23 is a large number, but 10, 12 are really active. And the other people are working on certain packages. But that's the point. 10 DDS said they are only DDS because David Amid exists. So in other words, if David Amid wouldn't exist, we would have 10 David Amid developers less. I think this is a really, really important point. And you can all make this happen if you try to reach out in your field and dragging more developers. Well, 40 DDS were in David Amid before David Amid started. So there are current answers. And the others had different reasons. And seven of these 10 developers above extended their activity to other fields. So they are not narrow-minded to medicine, but they are doing other useful stuff. And even some of them left the David Amid project and do other things which are very nice and it's fine. So only eight of them actually are active in David Amid and the other doing something else. So actually because David Amid exists, a very, very leaf topic in the David Amid universe may have 10 developers more. This is when available per year. That's not bad. So who's a David Amid team? Well, don't mind these wet lines. But mind here, we have a solid basis, right? These are a lot of people, these are the most active uploaders of David Amid. And we are not losing people. That's the point. If you see graphs from other teams, they are losing people, they are empty spaces here. And that's the point. We are trying to get the active people sticking to us. Same thing, David Amid user-mating list. You see, well, in the first years I was alone, basically alone. This was a hard time, I had fun as well, but now it works, people, yeah. You see? And please mind the difference in 2011. I'll just go back. In 2011, something happened. And I will ask you what happened. That's my task for you. This is discussion on the developer-mating list. This was, I said in the first years, I was alone by having two mailing lists. And then in 2006, we started the developer-mating list. Also here, basically the same pattern, but here's some who says, well, he's basically involved in another project. But the overall pattern is we don't lose people. This is about the bug hunters. People who are hunting for bugs and fixing bugs in Debian-made packages. And also here, 2011, okay. 2011 again, these are the commits to Debian-made VCS. And same thing, since 2011, more people joined and we are not really losing people. Some have some other things, but okay. So we have somehow a team. For me, a team is waking up in the morning and realizing that somebody else has solved your problem for yesterday. And it's fun if you observe this. It's actually, I was waking up, looking for my, oh, this package was uploaded, this was buggy. That's great. Yeah. We have even a Honored member of Debian-made team. Gregor, he's fixing our bugs. It's not member of the team, but that's great. So, well, the answer was here. What I wanted to know from you is what happened 2011? We started meeting. We started meeting each other. So in these charts, you have seen that this increase always started in 2011. Yes, it was some biologists reached out to other biologists, not only Debian developers, and we did the sprints. It was only in the biology branch. Well, it's the strongest branch. We should do it in the other branches as well. I hadn't the energy to organize this. And the ratio between Debian developers and software developers and users was about one to five. Christian, you were also there and this one knows not so many developers, but other people and this is our way to reach out and find new blood and it really worked. So closely working together with users, we have just one new people and it was one developer-passprint. The thing is that if you are a clever person, can code, use Debian. There is some block, some wall. You think, I really like Debian, but I can really join this project. Amazing. I was never aware that I can join Debian. And this is reaching out to people and explain them, well, it's easy. Come to our team and if you have people for different work fields, all these could reach out to their specific developers and users and they would join Debian. I promise you, they can do it. And by the way, thanks for supporting this print, which I think I have proven that they are successful. And here you can also see that it's successful. This is a graph of this direction. You see the meta package as a time and you see we have done this print in the biology field and you have also the rise here and the biology development and the other development straight, okay? But these prints have definitely an effect. We have not so much in practice management, but there is not so much software to package. There are some packages who claim they could manage a medical practice, but well, we are working on this. We have some other targets, but this is the development and you see in all fields, we get some more. This is nice. This is what I wanted to tell you. So I would call this kind of a healthy growth and after some stagnation in the beginning, we had some growth in all means in discussing, in providing codes, in adding new packages, in getting developers, but some kind of warning. The thing that we get more developers it doesn't mean that it's less work for you. You have also more work. You have seen also increased for me, but as I said, the work shifts to different fields. It's nice. I like this. Yes, but it finally solves the problem of manpower in Debian. We always claim, oh, we are no people. We have no available. Yeah, but you have to reach out for these people. It's an active process and I'm actively working for this. So my strategy was make the blends to even more attractive to make Debian made more attractive to users and developers and also make blends more attractive for teams inside Debian. So I'm, for instance, I remember clearly two years ago in Banja Luka, I was sitting in the Debian games, both sitting there and just preparing all the blend stuff. I was finished with this web page and everything on the end of the talk. And I showed them, well, you just need your, specific knowledge, blends are not only fun, you need knowledge, how to categorize, how to, what packages are there. And Debian and I have, don't have the overview. This knowledge needs to come into the task page. This is probably in the, in the DevTex, hopefully. Hopefully also for the new package, I don't know. We want to connect this to be even more efficient, but people, and that's true, regarding working on the blends as work. It is work, you, somebody needs to do it, but I think I have given you proof that it was doing this work to get more manpower. Well, and that's was my talk. You will find Google for Andreas Tiller talks and you find this link and yes, this PDF is online and you can click on it. You can now ask question or I can show you more on this task sentinel or whatever. Not so many questions. Oh, there's a question. Thanks a lot. Andreas, would it make sense to, to have a network infrastructure for other blends, like the one in Debian Edo, like the one that Andreas Mond is preparing in Debian LAN project? Would it make sense to have that as an underlay so that you not only can set up your own computer, but your own network? Yeah, I like the work of Debian Edo. I notice work. The thing is, somebody has to replace this Debian Edo by blend. And then it could work for any blend. Well, actually it's happening in Debian LAN. So you could combine a blend with Debian LAN and then you have an administrative infrastructure for a little office, like a medical office or like a guest consultancy. Yes, it makes definitely sense. And I will also be asked quite frequently, is there some DVD from Debian made? No, there isn't, because it is completely in the DVD set, but people are asking for it frequently. So I would be really, really happy if somebody would say, well, I will do it. It's not that hard. You can, you could use Debian Live, put all the made dash, dash, dash Asterix packages on it and you are done. So it's easy, but somebody has to do the work. Okay, thanks. Say we want to either start a blend from scratch or to revive a blend from your experience. What would be the best starting point, maybe creating some web page somehow to attract people, maybe to find people already doing stuff in Debian or what would, from your experience, what do you think would work? Well, my idea would be just, as I did for the games, I copied the Debian Junior, just copy some existing one and exchange the names of the dependencies and you get, after five minutes, these task pages and you can show people what you have. You get it in five minutes. It's automatically rendered, so it's updated. What's that package? Which package is that? Well, it's actually not packaged, but it's running on Elliot. Okay, right. And there's a code for this, it's in Git. So it is in the blend, skits and so you can, it's a task, in a Python script, you call it Python with the task name and then it generates the pages. It's generated from UDD and so it's easy. If you want to do something, just ask me and I will do it for you in a short time. Okay, that's a question on IOC from Petter. Petter Rionson, he's the- Greetings to Petter. Yeah, greetings. Oh, I have to type it. No, he's watching us, actually. So I'll type it anyway. So his question is to Andreas, would it make sense to use the task system for more for blends? So in terms of, if I understand him right and Petter, please ping me where I see if I don't. Well, he says extra information. Wunder provides a lot of task cell tasks while Debian limits the number. Should every blend show up in the task cell selection menu? Yes, I know, yeah. In the script which creates the meta packages is also a task cell control file included. It's, as I said, it's now overworked and so. So it's included. And I would really, really like this. There is an ancient bug report which is closed, won't fix also, which in the installer should say we have Debian ADU, Debian made, Debian games, Debian whatever. And I would really like to see this happen. Yes. Clear answer. Time over. So, yeah, I'm here for the WillDepCon if you can approach me and ask me some other. We can create a blend together in half an hour. Okay, thank you everybody. And I think we have a whole day of blends and buffs and so you have more opportunities to talk with Andreas. Thank you.