 Is there a trend for the project, or a weakness of the project, or whatever they would like to discuss? Maybe you could copy the, for example, all the strings, put them at the beginning of the document, and review them together, and put the other ones. I just read it. Yeah, so the things I identified back then in 2014, where the strong shared common ideas and goals contributed, so we have clearly a project where everybody, well, mostly everybody agrees about what we are aiming at, widespread agreement with foundational documents and procedures. Everybody cares about Debian's and likes it. If you talk to the outside world, I mean, outside the conf, lots of people are really big fans of Debian, even when they're not shooting it. We have a large active community of volunteers, and we value independence quite highly, and we don't compromise with it. So the other ones, I'll just continue reading, but Neil pointed the social contract as a trend. I think this kind of is included in what's line four on foundation documents. Neil isn't here, so he cannot comment, and maybe you can comment on yours. Oh, mine were. So there was independence, and I scratched it off because it was in your list already. There is, so let me show you. History of the project, we've been doing this for over two decades now. We have committed a lot of experience and showed that our work is ruined, there's a few software in our business community. It's to make sure that the structure is still working, having an organization with the good processes, and having, I mean, trying to avoid organizational blockers inside the project. The large community, so that's obvious. Largest number of derivatives. Someone is editing. It's a bit complicated with line six, but yeah. Line six, yeah, sure. Largest packages are repositories, and philosophy of technical excellence and commitment to free software. I don't think it's necessary to command them. They are simple enough. So what's missing? Microphone, wait for the microphone, I've had several people asking for audio during the book. I was just, Debian has a strong reputation for stability, and it doesn't seem to be captured up there. I think a strength of the Debian project is also it's non-binding to corporations. Independence. Independence, okay. Yeah, it was listed already. Actually, you like repeating things, right? That's great, isn't it? Yeah, it's just a way to get it in my mind. It's still early morning, no? Is that a sub-item of independence? Maybe we could detail in which way we are independent. We are also financially independent, not only about our goals, but in our philosophy and our finances. I was going to say, we have a huge number of very dedicated developers making it better, but there is large and active community, which I just saw now, maybe dedicated, put that in there or something like that, because it's amazing what we do, I mean, seriously. Go to the opportunities. I think it's easier for the other... So these are the ones identified back then. The first one is the fact that we are one of the few remaining large community projects. I would say true community projects for many other free software projects that are kind of polluted by companies. I don't see this that much present in Debian. We have approved many developers employed by companies to work on Debian, but still they mainly contribute... It's called them. Oh, sorry, I don't have the right resolution. They contribute to Debian with their Debian heart, and not with their, say, collaborator heart or whatever, or the company can put our HP or whatever. The second one was the general post-no-don OCT towards some big corporate players. We are clearly seen as being independent for all those big players. And CloudCell is here, is that the question? I think it was mine. Nils, I didn't copy them, because it's only about things and weaknesses that he didn't do. Nils. Cloud service is about... So, yeah. Zach made a talk at DevCon 14, if I recall correctly, about what is the definition of free services, cloud services. And he made a point about all the computing moving from workstations to the cloud. And the question was about how to make sure that people still use free things there, what is about free services, and also how to make sure that we are still relevant, like for example. So there is the on-cloud service. I'm not saying particularly that we should package it or something, it's just something that exists. It's not packaged today in Debian and got removed for some reasons. So somehow we are not enabling our users to easily install such software. And they may use another instance publicly available for free. But we don't know how it's installed, how is their data in the service private or whatever. So the subject was about making sure that we are working on packaging those services. Also about making sure that Debian is easy to install on the cloud platforms. So for now we still don't have, maybe we fixed it a bit, but not exactly. We don't have installers officially promoted by the project for like Amazon, Azure, etc. But are installers needed for Amazon or Azure? You just take image like AMI or some Google Cloud image or Azure image, like image of existing already installed Debian installation and you just use it. About Amazon for example, I know that there are multiple AMIs. So which one do you pick? Which one has been promoted by the project? We have Debian images. I'm not sure whether they are officially blessed, but they are named Debian with Debian Swir. On our download page, we only advertise Azure images and not the cloud things yet. Yeah, they are on the wiki. Okay, so when I'm using Debian on EC2 on Amazon, I usually create new instance and go to marketplace, put Debian and just start this Debian image. But at the same time, I was today at the Azure presentation and they collaborate, they are doing something really interesting like preparing every day, building new Debian images for a stretch, for Weezy and for Jesse, which is not the case for example for EC2. For EC2, we only get new images when we release new sub-release, for example 8.5 or something like this. So this means that we, and also we don't get relevant testing images in EC2, which means that working with testing some of the new packages on Amazon is quite different, quite difficult. I'm not sure what's the case with Google. Yeah, I think we agree on the technical status. My main action was about this one. It's about advertising, what it's being done, and make sure it's on the official download page and not on some wiki page. Okay, there will be both for the cloud in two hours. I don't know, Steve is working on that with his CD image hat to make sure that the cloud OpenStack images are being built by the official infrastructure. Yeah, that's also the question. I'm not sure what's the current status about that, but at some point it was quite, quite difficult discussion about what does Debian image mean on such public clouds. What are you allowed to include in the image by default, even if it's not in stable release? I think there is a page now, which gets the criteria. Yeah, but it's easy to apply to each and every public cloud we have images for. I'm not so sure. For now we don't advertise public, I mean official images. Maybe it's applied, but we don't know yet. Maybe I should put the link about that. Maybe just go on because the next item is yours as well. Yeah, the other one, it's not only about French-Arabic speaking countries actually. It's about also Asia, users from Africa or Latin countries, etc. Make sure that for example we go there, make some talks. I mean the project can fund that. So we have some representative, we have our project presented in conferences there. I know that a few developers go to First Asia, which is great, and we should be doing that for other conferences. It's a promotion effort basically. What do you see as the main blocure for improving? Well going to conferences is nice, but it seems to me that at least when I got involved, the main way people got involved was because one of their friends was involved in Debian. So it's quite hard to just bootstrap a new geographical location just because you have nothing there. But at least when you go there you have the opportunity to meet people, so you can encourage them to talk to you even by mail later, mentor them. Some comments off the internet. Jason's been making quite a few posts. Fortunately it doesn't seem we're streaming very well at the moment, so it is being recorded. So these may or may not have already been on the agenda. Continuing the definition of free software, promoting its work within Debian, retaking Sax Talk in Debian, DC 17, Debian in the Dark Ages of free software. We've covered most of that. Considering the transform of the official website, Debian.org, with less textual interface to generate more visual appeal and thereby facilitating access and read content to the general public and potential collaborators. Trying to attract more users, new collaborators and developers through the above two points as well as other possible activities. Please communicate in an opportunity. Yep, I think we've covered what's being asked for. Could someone with access to IAC please take a few notes about what was said by IAC? Rafael, maybe? Debian has a middleman. Basically we are the interface between upstream and users. There are moves to... with application building solution where basically we would no longer be needed. I think it makes sense for us to have our say in this move and maybe embrace it rather than refuse to accept it and continue to keep our influence on the ecosystem there. I took this example. There are other similar cases that we could find. We have at least our upstream guide which is sort of our official recommendations for upstream developers. So maybe it's not well enough advertising as well. Thank you. So I try to identify actions to do and to act on the opportunities like doing some advertisement work for the cloud services or images, etc. Do we have ideas for the other items on the list? So for example, for the website, I don't know the status of the WWW team. Maybe Paul knows about it? It's keeping up with bug reports and stuff but not really bigger things. So basically nobody is blocking it. We are just looking for volunteers to do the work, like converting the CVS stuff to Git, which has been on the to the list for years but nobody was... I do know that the designer that we recruited this year for DebConf, one of the first things she said is, ooh, the Debian website really needs some work, which she has had no time because we've had her working on DebConf but she might be recruited. We did have a slight problem with getting her to do free software tools so there's some work there but it's possible and she clearly has skills. About new contributors from countries outside North America and Europe, maybe that's something that should be more widely announced in a sponsorship program for developers, for DDs to go to conferences elsewhere to give talks about Debian tutorials about Debian packaging. That would be a good use of Debian money. Do we have enough manpower if we would get, I don't know, a few dozens of contributors, which would need to be mentored and guided at least initially? It's a luxury problem to solve the day we have thousands of contributors. So I don't worry, we will find enough work for them. But sometimes the problem is making sure that people are in touch. I mean, that's the goal of Ashish's booth, where you have the mentors saying what they are working on in Debian, the mentee saying what are their interests in Debian and it helps to make people in touch to start working on it. And they even manage to find new volunteers for local teams, so it's really working in minutes. So that group of persons, so maybe should... There was the idea of doing a welcome team actually in Debian which could help people start doing things in Debian. Sometimes people don't know who to contact to act on some area of the project. So if you know already those guys and if you know Debian enough, you could do this first step. Related to that, it might be more for the weaknesses. I think one problematic external impression of how you can contribute to Debian is that you absolutely need to be a Debian developer. I think we still haven't managed to fix that. And people quite often come and say, well, but yeah, you know I'm not a DD, so dot dot dot something excuses. But it's mostly our fault too, probably. Can I take issue with that one? Can I take issue with that one? I've not been a DD. I'm now a DDU, not uploading, but that only happened a week after last year's DebConf. And that was only because people were sort of saying, look, it's about time you actually went through the process. I've been involved with the project for 10 plus years. No real issues whatsoever. We're contributing without being a DD. It's purely an excuse not to do something. I don't think we disagree. Okay. I heard it as slightly differently, so my apologies. Well, I mean, I'm not saying it doesn't work. I'm saying there is an external perception that you need to have these ticks in your profile to be able to contribute. Maybe my impression of this external perception is wrong, but I'm glad it works. It works fine. There's plenty of people around here today. There's a lot of the local team on our DDs. I've been doing an awful lot on this because they've been not right, Tim, but he's been suggested by friends and people that have been working on it that this is really cool to work on. And becoming a DD will be further down that line, but it's not necessarily seen as a stopping point. I just want to comment on other opportunities. How should we move to the weaknesses? Yeah, so just for this one highlighted, for the apps, I suggested maybe it could be part of the roadmap because it's a wide archive change, and it fits there. Then we only need a driver for the idea. What about the Libre CPU architectures? Who added that? I added that. The question is actually what, from Debian, can we do to... Well, we can support them and create a port and so forth. Mostly that's referring to RISC-V, but there's also the SuperH architecture that the patents have expired in there, reviving it as an open source architecture. Yeah. The problem with RISC-V at the moment is the patches aren't upstream yet and they're still revising the ABIs and stuff like that. Okay. And this one, the Linux desktop? Who's about you? No, it was Wodix. Yeah, it was Wodix desktop. Was it you? Well, we haven't yet seen that year, so it's still an opportunity. Oh, okay. It's never coming. Just so you know, think about mobile instead. What's actually happening is the desktop is going away entirely. So don't worry about the year of Linux desktop. Who cares anymore? Do worry about mobile though. Okay. Okay, let's move to weaknesses. I find we're moving to weaknesses. I already copied the text. Yeah, so some of the things that were identified in the past is dispersion of manpower, lack of manpower on core things, lack of interest in contributors, common technical paths, fragmentation, lack of common technical practices in some areas. In fact, the packaging is difficult. This limits the use of packages, standard interface, standard software distribution mechanism. It's difficult to get started as a contributor, lack of sponsor requirements. Disconnection from upstream for many packages. Who is that? Odyx, again? You can comment on that if you want. I mean, I think we have a TC chair now who's under 35. I mean, it didn't happen since, I don't know, maybe 10 years. So it's a good sign that we have new younger people in the project. It's kind of the usual thing is that it's common for us to think that we're getting old and we're maybe not looking correctly and we might just be missing that the young guard is actually coming and there are actually people that are young that contribute. But I still think the first part of the sentence that the age average is growing stands and that might be a weakness. I mean, it used to be a project for the 20-ish people and now it's, yeah, we have a bar that closes at 11 and we are happy with it. Speaking from another open source project that went through this same process, I can tell you it is a problem. If you don't welcome the new generations, it means in another 10 years you're dead. But this actually ties into the earlier thing. We're talking about small events in regional areas. It's not even as big as a mini-Debconf, even smaller. Ashish did something awesome which was open hatch comes to campus. He would connect with local user groups and run. It doesn't take even, you know, maybe if one DD could go or even just someone who kind of knows a little bit about Debian could go. You can spin up these very small university events that get university students excited about the project. And that's probably the best way to get very young people involved. I got also in touch with my academic background in the past with a professor teaching teachers who wanted to do a short internship and we could have a program within Debian to have people from, I don't know, colleges or whatever to work on small tasks and it could be a way to get in. Okay, talking about universities. I remember when I was student we had some Linux user group locally. Now when I'm speaking with people who still keep up with my former university, now it's .NET user group, some mobile programming, some gaming. But I'm not sure if Linux user group even exists at my former faculty. So that might be the sign of the times changing and that might be also one of the reasons that we don't have new blogs because my first contact with Linux was at the university and we were exchanging some ideas, we were discussing about distros flaming there. But I think we also have to update our speech when we are trying to reach out to new users and communities because we always start by saying I could do, for example, a packaging tutorial but people are not interested in packaging and we have to update our speech to find new interesting tasks to do in the project which would be appealing and interesting for young people. So if there aren't Linux users groups on campuses anymore, is it because it's already a done thing and people are using it in the engineering programs or is it not? I'm old enough to have a son in college and connecting his computer to the internet you needed Windows or Macintosh. If you wanted to run Linux you had to run your own network. But I know a lot of engineering programs are already using Linux. So it's not only about the users groups it's do we have attention of the faculty and incorporating Linux into the curriculum? I think that we are at the bottom of the curve currently because I think that traditional operating systems courses are becoming a bit less relevant because there's not so much of a need for people with such skills on the market, at least probably less than 10 years ago because everybody uses Linux and it just works. And at the same time there are not so many people who identify the need for DevOps or managing large infrastructure skills which are more a mix of system programming and assistive meaning and so nobody is really teaching about those skills at the moment. We can hope that this will change just because many large companies run really large infrastructures and there's a need for people to do that and clearly that's something that is quite hard to find on the job market today. Okay, so if I may play with your idea. So it seems like we have two problems. One is the lack of role models of somebody whom to aspire like faculty using Linux and showing that you can do cool things and also lack of view like Linux now is seen as operating system yet another operating system like Mac or Windows or whatever and it mostly works. So it's not seen as students and new fresh blood is not feeling the need to tinker with it and from this tinkering we had new contributors. So we would need to find some idea to freshen this tinkering. I don't know maybe returning a little bit to the cloud maybe show them how to manage some few instances on the cloud. I recently got email from Google that they are providing some credits to run some Google cloud in academia so maybe get from this angle because when it comes to the new packaging I'm not sure if that's really pressing issue. We have over 50,000 packages and in my opinion more problematic now are orphaned or not really well maintained packages. Staying with your academia theme. Certainly in the UK we're finding a lot of the colleges are demanding submission of tasks and reports in a particular system in a particular format and they're all demanding the documenting word format or whatever and there seems to be a lot of effort has gone in to prevent it being submitted from anything other than an Office 365 generated device. Almost as if I were being carefully pushed out deliberately. The universities don't want to take submission of an electronic document from MOA or LibreOffice. Even for PDFs? They don't accept other things like PDFs? No, they've now got their plagiarism. Certainly in the two universities I've had contact with recently demanded document in a word format. They've run it through their anti-plagiarism script. So it's expecting something in particular. It then goes and processes it. So what you've actually got there is freedom is constrained intentionally to provide a service which you must go down a particular route. So they're expecting the students to be running on on Windows Box or running on Mac. So it's a tooling problem on the universities and since their budgets are getting smaller and smaller. Yeah, budgets being constrained, but of course freedom is also being constrained as a result. Still I think this issue is a bit different from the one before. The fact that people cannot use Linux-based systems at universities is one problem. The fact that the university isn't teaching that much about the importance of Linux-based systems or Unix systems is another issue. The one you mentioned applies to everybody at university. The second one applies to computer science students. And in terms of recruiting contributors, we mostly care about the second group of people and also, yeah, it's still... If it's going to grow, we're going to need to appeal to more than just our potential developers. We're going to have to appeal to our end users and our end users need to be everybody else. It's a larger group of people to appeal to. Can Raphael comment on the... before the last one? And clear decision processes or... Oh, sorry, really? Yeah, Madoc, sorry. Kind of assumed that it's a little bit self-explanatory, especially the first one. But I guess it's the huge picture or circle around how do we progress, how do we make progress in the project. And often there come some controversial decisions that need to be made. And some teams are really good at doing that now, communicating up front and then making the decision and just moving forward with it. But it's not clear how that works, how it's expected from everyone in the project. And for instance, if I wanted to drive a certain change in the project, it's not clear how I even would get to the point of identifying who can make the decision without going to a GR. And I guess it ties in with the second point that I made there, the reluctance towards entrepreneurial initiatives, which may well fail. I have this feeling that in Debian we're all very protective of what we have, of our brand, of our project, of our pride and everything. And yet at the same time, if we don't take risks and try some stuff out, we're probably not going to be able to advance a whole lot. I've noticed a very strong decline in innovation capacity in the last decade in Debian and maybe that's partly related here. So sort of embracing this sentiment that it's okay to try something and if it fails, then we'll just pick up the pieces and keep going rather than fearing that it might negatively affect the brand Debian. Because if we're perceived to the outside as an innovative project that tries stuff, I think that's going to be a better image to Debian even if some of them fail than if it's a sort of stagnant project. So is it about promotion of individual projects? Well, pretty much anything. I was thinking when I wrote this, I was thinking about the partners program, where some of this entrepreneurial freedom would be very helpful to have, but I'm sure that it applies to many other things, you know, events. Probably less so the technical stuff. I mean, you're not going to be like entrepreneurial about technical innovation in some ways. We do have processes there, but there's a lot more around. Take for instance, a website design. I mean, one of the reasons why we do have an updated website, it looks better than 10 years ago, but there have been a couple of people trying to make it more of a friendly, responsive design and so on and so forth. And while I'm sure that a lot of people will be in favor of that, the sentiment in the project, if I had to describe it like that, is more like very, very skeptical of the whole thing. And so therefore when somebody approaches this, they go like, I don't really want to touch this project even with a broomstick because I'm afraid that I'm going to get backlash. Whereas what we should have is a community where if somebody says, I want to work on the website, everybody goes like, yeah, that's awesome. We're looking forward to having your designs and then we'll provide feedback and so on. Rather than you need to... You are very conservative with what works already. And that's also good. I'm not saying that we should start failing on everything, right? But it's true that when we have new initiatives, we tend to ask too much questions that eventually discourage the person who was volunteering for it. Instead of, for example, making a test bed so they can work on and show it to the people, they... But failure and experimentation should also follow with some post mortem or assuming that it failed or post some roadmap for doing this. For example, we have DebianDroid, which was the result of Google Summer of Coast two or three years ago. But students finished this and this application was not touched in three years already. So it's now obsolete. It doesn't display correctly some of the bugs. And it has only below 100 users. So when you put Debian into Google Play Store, it's one of first hits and it's not really good. So should we try to do something with it, like rewrite it with modern Android guidelines or I guess this goes a little bit with our social media presence. Because in my opinion, if it would work, it could be a good first point for somebody, some Debian user to subscribe to some bugs to show what are the versions of some interesting packages. And at the same time, a little bit less than our burden on web page. Because if somebody can look it on mobile application, then it's more modern, more hip or something. Any other comments? Should I just pull it down a little bit? It's okay. So innovation capacity decline. So looking at the first thread or second is the end of the desktop and we already talked about cloud images. Under opportunities, I think we should talk about Debian on devices. As a real opportunity. If you look at Beagle board, you look at Raspberry Pi, they're both using Debian today. I think that's an opportunity. If you look at the display case out front, you see all those robots. I'm willing to bet they're running something, some type of embedded Linux. So if you really think the desktop is diminishing, you have two fronts to focus on, the cloud and devices. So we should really be looking at both of them. There are a few minutes left. Someone wants to comment on some of the threads. Inspiration of thread art. System B. Rafael, do you want to comment about that? They have insider information. What I was discussing was I read that guy who told me there are many parts of the lower stack GCC and also other stack in GNOME and everything. And if threads were to die, we would have many problems because Linux would work much worse. He mentioned, for example, that well, that does a lot of laptop testing and many of the drivers actually do work because Red Hat engineers fixed them. I don't know how much truth it is, but I think it's believable that Red Hat's involvement there is important. And well, those are sort of not very realistic threads, but still it may be nice to have more devianish people involved in the lower stack level where Red Hat is heavily involved so that if Red Hat disappears, we have people left. I want to comment on the containers as a solution to deploy applications and services. Sorry, what? Which is identified as a threat, but we could also look at it as an opportunity. At work, we are maintaining a devian derivative internally based on Stable. So we have the old Stable releases as well as testing. Many users are asking for a back port of some applications from release to an earlier release. What we found simpler is to have some process and programs to easily get applications run in containers from newer Stable releases. So it's technically very easy. You just have to install a chute, install the packages there and make some scripts to run the application. So it's easy to do and it's a way to say to people, okay, so if you manage to package it in testing, you are not forced to wait until it's released as Stable to use it. You can use it from today. We don't have such tools yet, I mean devian. So if we... one minute, okay. I don't have anything to say more about that. Just work on some tools around containers. It's a new technology and we didn't take advantage of it. We are only looking at it as a threat. It could be an opportunity to make things simpler to distribute. I think that generally we are not really embracing all the new ways to distribute software up to their potential. One thing that is clearly changing is that people want to use different versions of software and just use backpots and everything. Backpots are not really a solution because most of the packages in devian don't have backpots. One idea I've been playing for a long time is trying to do automatic back porting. I wonder how many of our packages would just work to feed them for the target distribution and provide them as untested stuff to users. Most of our users just do that from source and we are out of time, so I'm stopping there. If you're interested in working on that, I'm interested in talking to you. Thank you. Thank you.