 Okay. Okay. Hey everybody. Welcome to our next talk about the relation between Debian and free software foundation. So please give a warm applause to John Sullivan. Thank you. So my name is John Sullivan. I'm the executive director at the Free Software Foundation. I work at the Boston office. I've been there since 2003, not coincidentally also the first year I became a Debian user. I've been the executive director just since 2011. And in my free time outside of that, I've been a Debian developer since 2010. I don't think contrary to what Lina said last year that I'm either crazy or on drugs, but I will leave that up to you to decide, and I will plead jet lag most likely. So unfortunately, as my FSEF responsibilities have increased, my time for other free software projects like Debian that I worked on in the past has decreased. So I have to confess that my packages are in a pretty terrible state right now, especially the crucial X word package, which I know you've all been missing. How do you do your Sunday crosswords without the X word package? I'm hoping to spend some time this week to get those things back in a better state and make some time to contribute meaningfully. I'm talking here as an FSEF representative, but as a Debian developer, I also really want to see a closer relationship between the two groups and between the two projects. It's something that's really important to me personally, but I think also really important to both the FSEF and Debian as projects that are uniquely committed to free software ideals. I really appreciate the chance to talk to you all about this today. This is a bigger group than I was expecting. I was surprised to be scheduled on the weekend with the emphasis on the open public day. I'm really honored by that and really appreciate you all being here. So related to that, can I just ask for a quick show of hands? How many people saw my session last year in Portland? Okay, so less than half. So I am going to cover some things that will be a bit repetitive from that last year, but I think it's necessary in order to set the context and make some of these points clear. And we have a, well, we have a bit less than an hour here. And my plan was to try to leave at least 15 minutes at the end for a discussion. I'm here to answer any questions or hear any feedback you have for the FSEF in general, in addition to the points that I cover in the presentation. But you can still interrupt me if you like as well along the way, if you have questions that are especially pertinent to something that I'm saying. So other than LibrePlanet, which the FSEF hosts every year in March, DebConf has quickly become my favorite event of the year because it is so committed to both the community and the production of free software. And that is very different from what seems to be the trend at other events that I attended throughout the year, which have a greater emphasis on figuring out how to support proprietary software using free software. So it really feels good to be here. And I am in awe, I would love for LibrePlanet to be able to travel around the world each year instead of always be located in Boston like DebConf does. But I have no idea how you guys pull that off year after year so well. You know, speaking of working together, maybe we can co-locate an event sometime that might be a fun thing to try and also, you know, my secret agenda to figure out exactly how you manage to do this so well. I hope that some of you, nevertheless, will make it to LibrePlanet in Boston next year. We've had several speakers from the Debian community there over years past. And it will be awesome to have more of a presence there. Also, you know, we have a bit of a friendly competition, and now we have to step it up since you've surpassed LibrePlanet in terms of attendance. We need to do a better job of getting more people there. So congratulations on the record setting registrations this year. It's awesome to see. This is actually only my third DebConf, but it's my second one in a row. So this is a nice opportunity to be able to check in and present some updates about some of the project possibilities as collaborations that I mentioned last year during the talk. But first thing I want to say is happy birthday. I couldn't celebrate too much last night because I had to be, you know, awake and coherent to present at what is 5 a.m. my time. But I did enjoy being there and the vegan cakes were delicious. Thank you for having those. We should also say happy birthday to Genome. We returned 18 on the same day yesterday. So can we say happy birthday, Genome? The FSF is also, well, first I want to say that Debian has been and continues to be an incredibly important project for free software. And nothing that we talk about in terms of the FSF's endorsement or the way that we see particularities of Debian should take away from that big picture view. The FSF has had many servers running Debian over the years with only main installed, of course, and several other distributions that we actively endorse and promote actually are derived from Debian and couldn't exist without them. So I want to thank you all for your hard work, for the quality of the work that you do, the technical quality and also for your commitment to your free software ideals and the principles that have driven Debian for those last 22 years. I think it's incredibly valuable and part of why I'm here is to help see what I can do to help in that work. So the FSF is also celebrating a milestone birthday this year. It's our 30th birthday and we are having a celebration that I hope some of you will be able to make it to. On October 3rd in Boston, we'll have both a mini user summit, user freedom summit during the day and then a party in the evening. And if you can't attend in person, I hope that you will check out. There will be local events around the world that people are already stepping up to organize. It'll be fun opportunities for people to get together and celebrate free software and what the free software movements achieved. So today I want to talk about two main themes. The first of which is the question of why the FSF, the delicate question of why the FSF doesn't fully endorse Debian and what the possibilities are for changing that situation. And then the second thing I want to talk about is what I think we can work together on in the meantime, even if the meantime turns out to be a really long time. I chose the bad punny title of the session, solve problems at the source, get it? It's FSF and good new tradition. Because I happen to think that if we were together, working together effectively on some of these specific projects, we might find ourselves solving the perceived disagreements by side effects and getting ourselves closer to the world that we all want to have. So if you look at our big picture goals between the FSF and Debian, they're suspiciously similar. So the free software foundation goal is to have all computer users to be able to do everything they need to do on any computer using only free software. My reading of the Debian goal is that we want to have a universal operating system which will remain 100% free and never have any non-free requirements. So I take universal to mean able to run on any computer, big or small, to be used by technical and non-technical users alike with no non-free dependencies. So to me, these visions look strikingly similar. In terms of distributions for the FSF, we don't have a single low and official distribution, but we have a list of distributions that we think are actively implementing the goal that we want to see in the world. These distributions have two key commitments. The first is to remove all non-free software and to not ship any non-free software and also to avoid steering users towards non-free software outside of the distribution. So last year at this time, the list of distributions that meet these criteria looked like this. It included a blag from left to right there, a blag, dragoire, dynabolic, canoe scents, musics, parabola, triscoll, which is what the FSF uses primarily on our workstations and servers, and which is based on Debian by way of Ubuntu, and Utsuto X, which was the first distribution that was endorsed under these criteria. And there has been a few updates to the list since last year. ProteinOS, which targets embedded devices but is also intended for use on laptop boot systems, was added. This is a distribution that can be used in the future. We hope to replace things like the Intel Management Engine firmware that runs on a separate CPU at boot time in the machine. You can actually, we hope, in the future, be able to flash that proprietary operating system with ProteinOS. We also added Libre CMC, which is currently targets wireless routers, but has the goal of working with other kinds of embedded systems as well. And this was a nice announcement because a couple of months after we announced the distribution endorsement, we were able to announce certification under our Respects to Freedom hardware program of a wireless router, a home wireless router, running this distribution. So I think this is awesome. You can now, from ThinkPang, when you're running a wireless router running 100% free software. I know a lot of us have been waiting a long time for that. There's always been something, either it's the proprietary blob for the wireless or the bootloader for the router. There's always been some non-free component, but all of those things have been replaced with free software. So these two distributions together created a new section on our list, actually, which is focused on small system distributions, ones that are targeted towards systems with little resources like routers and small CPUs inside the machine. We also added the Geek system distribution, the last one there. Some of you might be interested in that. It has a declarative operating system configuration and has other kind of interesting features like the ability to do transactional upgrades and rollbacks on a system and have per user installations and packages without giving out root access for pseudo. So it's a pretty cool thing, especially for all of you that are interested in functional programming. So why isn't Debian on the list? So Debian is actually in a special category that's not reflected on the list. And we've been looking for ways to do this, but haven't found a satisfactory one yet. So it's the only distribution that meets the first criterion to not include any non-free software and to remove expeditiously any non-free software that's discovered in the distribution that's not already on the list. And the FSF fully acknowledges this and appreciates the work that's been done in the official Debian distribution to bring an all-free distribution to users. So that's something that we want to highlight. When Squeeze made the decision to remove the non-free software from the kernel, that final remaining bit, we put out an announcement celebrating that achievement and highlighting Debian for doing that. And there is no other commonly used major distribution that's in that category. Every other distribution that's considered common among users contains non-free software in its official distribution. So that is something that sets Debian apart. And I know that this message is not always clear. In particular, I know that some well-meaning FSF supporters have a habit of saying that the FSF thinks Debian is non-free. And that's not true. And you can now cite this presentation as evidence that that is not true. We say very clearly that Debian main is all-free software and upholds commitments just as good as what the FSF expects of ourselves in its official distribution. So in Debian's case, the lack of endorsement from us is primarily because of the relationship between official Debian and unofficial Debian, the non-free and contrib repository. And that relationship to us seems too close for our comfort. There are spots in the Debian infrastructure where those sections, even though technically separate, are integrated very closely with main. So for example, in package searching, in recommends and suggests fields within packages that are displayed to users. So even though I think in Debian we have an idea that these are separate, that's not always as clear to users on the outside and they can end up being sometimes inadvertently or sometimes just led to install non-free components on top of the official distribution. There are also some cases of individual software packages in main after installation, leading users and recommending users to install non-free software. An example of that from last year in the presentation was Ice Weasel where the add-ons menu in the browser had a very prominent recommendation to install Evernote, which was a proprietary add-on and did not say that it was proprietary in that view inside Ice Weasel. So there are cases of this where the Debian packaging isn't recommending any non-free software, but immediately upon running the software and clicking on an option within the software, the user gets shown non-free recommendations. So other distributions have been struggling with this problem as well and we're having to develop separate extension repositories in some cases for different programs in order to extend that commitment to free software a little bit further. I'll talk about some of that toward the end, I hope. So I think these issues are important to discuss within Debian regardless of what the FSF thinks. I think that the social contract language that Debian remains 100% free and has no non-free requirements has tension with the idea that you wanna support people who create or use non-free works in Debian that you'll support the use of non-free works and provide infrastructure for non-free packages. There's a lot of play there that's tough to figure out and I think for Debian itself those are important topics to be discussing. I'm happy to see there's another presentation, at least one more here that plans on tackling those issues. So I know that some people see the FSF standards in these areas as either a kind of censorship or as anti-user in some way by trying to kind of hide software from users or pretend like it doesn't exist. But I think that Debian itself has many policies in place which are designed to steer users one way or another when it comes to the software that they use. It seems to be a primary reason why the contrib section exists in the archive, even if packages are free, if their only purpose is to point to non-free packages or install them, they go in a separate place outside of Maine. Or there's limits on ways packages in Maine can refer to packages in non-free, especially when it comes to dependencies. And I know that since the last EPCOMPT there was a technical committee decision that listing a non-free alternate dependency in a package in Maine is acceptable. And I was a little bit uncomfortable with that decision but I also saw clearly that it is couched in terms of caution, you know, the actual presentation of the decision said that package maintainers should be very careful about using this because of the way it can lead to use of non-free and also that the Maine dependency is still a requirement. You can't have just a dependency on non-free. So I don't think, you know, even though the decision is not consistent with what the FSF is looking for and the distributions we endorse, I think it also didn't, you know, change much about the things that we already have to talk about. I think that the act of making a distribution is by definition act of steering users one way or another. And that's why in some senses users choose a particular distribution because they like the way that it steers them and the kinds of software that it presents to them. So I don't think that Debian is actually uncomfortable with this idea. I think it's a question of degree and of framing. And framing isn't just about semantics. It's about the questions that are posed and the way that they're answered. You know, are we talking about it in terms of censorship or are we talking about in terms of supporting users who want to just one time choose all free software and never be presented with a choice to install something non-free again? You know, so the framing of that question matters and I think it's important that we talk about these things in specifics. You know, another example is security. Debian will remove packages that have security bugs that don't get fixed after a long time even though users may complain loudly about that because they want to continue using the software. So Debian makes a judgment about what kinds of software it wants its distribution to be associated with and then sets up practices and policies to facilitate that. So in the end, I don't think broad generalizations about steering, recommending, suggesting are really that helpful. It's a gray area. It's not a bright line. And I think that we need to focus on the specifics of where we choose to draw that line. But the quickest way out of all this is to just make the vast majority of hardware or all hardware, if possible, be supported by free software and to have free applications, free versions of all applications that users might be said to require under the social contract. All right, if we have those things, then we don't have to worry about these questions. So back in at the end of July, the FSF board had its first ever strategic planning meeting and the most important five year goals that were discussed at that meeting I think are actually goals that are the vast majority of Debian developers probably also agree with. So I wanted to share some of those. So we want there to be readily available in contemporary hardware products that work with and contain all free software, in particular laptops and tablets, but other devices as well. It's awesome that a Libreboot X200 was or it wasn't given away this morning or it's going to be given away. Did somebody in here win it? Congratulations. I was really happy that we were able to certify that laptop under our Respects to Freedom hardware program earlier this year. We are, despite no cooperation from manufacturers, still catching up in terms of the modernness of the laptops that we've been able to endorse. I was actually using an X200 anyway at the time that the endorsement came out. So I know it's not as contemporary as a lot of people want and that's why we still have this as a goal but I see us making some good progress there and I'm really happy whoever donated that and made that a part of the conference this morning. We also want to end regulatory mandates for proprietary software. These exist in all kinds of places, from the FCC to the IRS, to accessing government documents of various kinds. All of these things create cultures around the world that are hostile to free software and therefore are hostile to Debian. You can't be an official Debian system user and interact satisfactorily with those systems. We want schools adopting free software and compatible hardware for students. And I think initiatives to promote free software in those contexts also leads to great opportunities for Debian. I assume that part of Debian's universal goal is related to education and schools and certainly any long-term play at getting more people to use an operating system involves getting people using it at a younger age and getting comfortable with it. We want the general public to see free software as their issue. Debian also wants the general public to be Debian users, to be free software users. I think many of us in the project don't like the reputation of Debian as an operating system that's specially suited for citizens or servers. We think that Debian is and can be a distribution that works for non-technical users as well. And I think that getting the general public to understand that and give it a try is something, is a goal that we very much share. So obviously, I said five-year goals, we're a little bit ambitious here. And these goals need to be fleshed out a lot, they need to be quantified and made more specific. But I wanted to present them, they're only a few weeks old here, but I wanted to present them as a way to show that I think on big picture issues, the F7 Debian have a lot of agreement and that professed disagreements have a lot more to do with things at the level of tactics and what's right in the current moment. But in the end, we agree on things like this. So I think we should collaborate more on projects that get us closer to these things. Last year I presented this list, and the first thing on the lists was the free software directory. So the FSF has been building in this directory of free software, individual free software programs. Initially it was supported by UNESCO and for many years it's been developed as a manually curated collection of free software that has no non-free dependencies and follows a basic level of good licensing practices. Does that sound like any other projects maybe that we're familiar with? It's a lot like Debian. The active packaging programs is a curate, active curation. Debian packages and developers become a little bit familiar with licensing. They have to be in order to make packages that get accepted. And of course, we all keep an eye out for things that don't follow good practices in those areas. So before a few years ago, the directory was maintained by just a few people. We had a full-time staff position to do the majority of the work along with a few great volunteers, but this approach never really scaled. So a few years ago we decided to convert it to semantic media Wiki, enabling more people to add entries and keep those entries up to date. So truthfully, it still isn't fully wickified. We're using the approved REVS extension, which means that you have to be granted some extra access in order for your changes to be immediately live, otherwise they have to be approved by somebody. But we very readily give out editor access once a person's demonstrated that they understand the basics of the criteria for inclusion in the directory and that they understand the process. A lot of that fault lies with us for not having better documentation, which we're working on. We have weekly IRC meetings every Friday in Pound FSF on FreeNote, where people have been getting together to add and update entries. We no longer have the full-time staff position, so these meetings are run by Joshua Gay, who's the head of our Licensing and Compliance Lab. It turns out to be a very appropriate match because the hardest, the biggest challenge in deciding whether a program should be added to the directory or not is figuring out whether it's actually free software and whether it has any non-free dependencies. I think as a lot of us know, from Debbie and the work upstream doesn't always make that very easy. So it's an area that requires that kind of licensing expertise, which by the way is part of the point of the directory, because telling people to use free software when it is actually not that easy to figure out whether a given program that you find on the internet is free software is asking a lot of somebody. We want to make that easier. So this process, new processes worked out very well and we've seen a significant increase in participation. But at the same time, in parallel, we've been working on an automated system, import all of Debbie and Main into the directory. And this was something that we had in mind from the beginning of the decision to convert it to semantic media wiki. And so from the beginning, we were looking at things like dev tags as the guiding light for how our semantic property system should work. We wanted to, first of all, we thought it was a good system and second of all, we wanted to preserve hopefully forward compatibility for importing entries from Debbie. So the idea here is to create entries in the directory for each appropriate package in Main and then also to be able to automatically update those packages with the system to reconcile those updates with any manual updates that have been made. And this whole process is made possible by the awesome system that Debbie and already has for structuring packaging information and making that predictable and regular within packages. One thing I have as a hope out of this project is that we'll be able to help with some of the initiatives that seek to improve that system even further and that we'll be able to help. And I think we already have a little bit with helping to make sure that packages consistently apply those standards. Because we find, since we're doing this in an automated fashion, we find hiccups and inconsistencies between packages. And I don't know if we've been perfect about this, but we definitely have a goal to file the appropriate bugs in those situations and patches to help fix it. So as part of an initial test, since my talk last year, we've already imported about 1,500 packages as a test run, which brings our total up to about 8,700 packages. And the next import, which may be happening as we speak right now, it was supposed to happen this morning. So there may be entries being added from Debbie and Main to the directory right now. This import will bring the total number up to 15,000. So it'll pull in another 7,000 or so from Debbie and Main. So I'm not the best person to talk about the technical details of this, like why aren't the all 43,000 packages already imported or what the challenges are. But if you wanna drop by one of these IRC meetings or find me, I can connect to you with the right people who can answer those questions. And of course, you can read the source of the importer script, which is in that Git repository hosted at Savannah there. And I'm sure they will be happy to accept any patches or feedback from people who, the actual script work was started by DAF, a Debian developer that a lot of you know, and has been continued by Rubin, who was the founder and the developer of Triscoll, also very familiar with Debian and Ubuntu packaging. So they have had some success, but they can always benefit from more eyes and advice from experienced folks. So why is this work important? Can't you just Google all the programs you need? A lot of people have been kind of, I've heard criticism of this project all on those lines before, but you know, actually increasingly, I have been feeling that the work of building this manually curated catalog is incredibly important for free software adoption. And in the long term, therefore, Debian adoption. So a big problem that we face with convincing users to switch to GNU Linux is to convince them to try a free application to begin with. They're not using GNU Linux, they're using Windows or OS X. So it's asking a lot of somebody to say, just change your operating system, trust me, everything you need is there. It really can help to have individual programs that run on multiple platforms that people can acclimate to. And then after they're using a couple of those or a few of those, they may feel a lot more ready to spend the time to try out a distribution. And then they like what they see, we hope, and they finish the process. But we face an obstacle with that, which is there's proprietary software companies out there telling people not to download anything from the internet that doesn't come from one of their official locations. And frankly, given the state of, you know, computer security and malware, that's not terrible advice. So it's kind of hard to be the ones fighting for, no, go ahead, install any program that offers itself to you for free on the internet. That's not a good approach. On top of that, you know, users are expecting software to be presented in the form of, you know, app stores and curated locations now. To the extent that we have those for free software in general, outside of distributions, they're mainly in directories that are curated for the purposes of advertising revenue and not really for the purpose of helping users. And then there's the fact that even for experienced users it can be hard to figure out what the license of a program is, like I mentioned. So we have a couple of goals here. One is that we want to use the directory as a canonical source of programs that are free and then provide a mark that people upstream can display to show that they're included in the free software directory and that can be propagated as a sign of trust for an application. Not necessarily like that it's an awesome application or that it's perfectly secure, but at least that it is a free application. The source code is available, follows some basic best practices. And then I also hope that it can turn into this sort of app store free software marketplace for users. And I think that Debian and other distributions have been, we're doing this app store thing long before any of these came along and been doing an awesome job at it. But that only kicks in once a user makes the leap to install it. So I think that it is in Debian's interests to help with efforts to encourage users to make those first efforts at trying out applications that are also present in Debian on the proprietary platforms that they're currently using. And I think that can be a way to get people comfortable with the idea of full GNU Linux distribution. People have talked about building other app store like features onto the directory, things like reviews, user ratings, demos, ways to donate directly to projects. I think there's a lot of potential here with something like this. And I look forward to seeing some of that materialized. I think in Debian terms, it can be a great source for ITP and RFP bugs looking for packages that are upstream free software programs that aren't yet packaged in Debian. I mean, that's the goal and we're going to be importing, hopefully packages from other distributions as well. So it can be a place to look for seeing what other distributions have packaged that maybe hasn't made it into Debian yet. So getting involved in the project, those are the basic ways. If there's enough interest in the Debian community, maybe this can become a joint project and some more official, higher profile capacity. But either way, if individuals just want to help out, that's awesome. And I want to definitely say thank you for doing the work to make this possible to begin with. This is a free software directory that only lists 5,000 packages before we started this. It's not nearly as good as one that will build on the work that you all did and list 40,000 packages. So the next thing that I had talked about last time was the situation with some upstream policy issues like Mozilla. This situation has become more complicated since last year. The Windows version of Firefox is now calling out and installing a proprietary Adobe DRM module the first time Firefox has run. They're not yet doing this for the GNU Linux version, but it seems to be in the cards. And I already mentioned the issue with the proprietary add-ons. So other programs do this add-on approach better. You know, Viber Office, for example, has a list of extensions. Its official list of extensions is all free software. So it's not as though this is some pipe dream. The majority of Firefox extensions are free software. You know, it's just, I think, something that we share that we want to make sure users see those things clearly marked and that Debian should be primarily leading users towards free software solutions for problems rather than the proprietary ones. So the main action that's happened on this over the last year was that as part of the free software directory, actually, if you look at the Ice Cat page in the directory, but this applies for Ice Weasel as well, you'll see a list of extensions that have been vetted and added. So there's a system within the directory to indicate that one program is an extension for another program. And so we're using that to build a list of all of the free extensions. I think this may become more urgent because of Mozilla's decision to only allow signed extensions in the official Mozilla repository. Signed extensions are great, but they're not allowing users to modify the keys in the browser, the accepted keys. So Firefox will only allow extensions that are signed by the official key. There's no interface for users to change. Well, there is interface for users to change that. It's called rebuild the pack, rebuild the software from source with the new keys in it. So I think there will be demands from other users to install signed extensions from people they trust. So various developers that they're familiar with, even those extensions may not be signed by the official Mozilla key. The next thing that it talks about was the database of hardware that's compatible with all free software. This, I was happy, we actually put out an announcement shortly after DebComp last year together. To say that we joined to help free software users find the hardware they need. So the FSF had been helping to build this database different from other databases of compatible hardware. Normally, those are presented just as a question of does it work with the kernel Linux or not. Irregardless of whether it's proprietary firmware or proprietary drivers required. So HNode takes a stricter approach and only lists hardware that works with a fully free system like Debi and Main. So previously we had said that users needed to be running and this information all is user submitted. We'd previously said that users needed to be running one of our endorsed distributions in order to contribute information in order to avoid inaccuracies and slip-ups from people thinking that something was supported by a non-free driver or by a free driver, not realizing they had a non-free driver installed. But we made the change to add Debi and Main specifically to the front page of hno.org as also a valid and recommended source of information for this. So I'd really love to work more with Debi and folks on this. There needs to be more information in the database. The site, the code, the power is the site itself. Needs a lot of help. Design could be better. There are translations, there could be more. There's lots of opportunities to help out with different skill sets there. And I think that I know there's been a lot of demand within the Debi community to how can we get the official distribution to work on more systems since we no longer have the proprietary firmware there? You know, how do we make it easier for users? So I think the first step is to make it easier for users to figure out what to buy in order to use Debi and Main. And I think that's something in the Debi and Project we could do more on. And in the FSF, we're also going to be working hard at doing more in that area. So if you'd like to be employed working on any of these projects, we are actually hiring for a deputy director position to work closely with me and the FSF staff and board on these issues and it does come with a star. If you are okay with living in Boston, which I'm told at least is the most European of American cities, it's, you know, our beer's not bad either. And short of that, I will be here until Sunday and we'll be happy to meet with people that are interested in talking, working on any of these projects that I mentioned or talking about other possibilities, other ideas. You know, I kind of lied about leaving a lot of time for questions, so please feel free to grab me and find me. And there is still the FSF collab discuss. Oh, I don't have it on there. On Alioth, there is an FSF collab discuss mailing list. That's not been very active, but it is a great space in the past, but it's a great space that we can use now to continue conversations about this. And I'm subscribed and will read anything that people post there. So the bottom line for me, for all of this, is that while the FSF is not promoting, in the short term, Debian as a distribution, we are working hard on projects that will help the official Debian distribution flourish. And I think when you look at our utopian goals, they're very compatible. And I think that's a very good reason we should be finding more ways to work together to make those goals a reality. I think that if we do that, we'll see the disagreements and perspective and even probably see them start to make way for more and more areas of agreement because we really do have a lot of them. So thank you. So if there are any questions, please line up to left and right side. Hi, we heard a lot about how Debian should change to be included in the FSF free software directory or free digital directory. What about the changes Debian would like to see in the GNU project or the FSF, like the GFDL being considered non-free by Debian in some variants? Yeah, so first of all, I do think this is a dialogue that needs to changes on both sides. And I'm sorry, I didn't say that explicitly. In particular, I think that if you look at the distributions that we've endorsed, they've been smaller projects and the criteria sort of work differently when you're talking about smaller projects versus projects the size of Debian with the number of users that Debian has. So I can see our criteria and where we draw the line in that gray area being influenced by the conversation with Debian as well. If I had very specific changes that I thought had to be made, I would provide those. But I do think that it's actually a conversation that needs to happen. And with regards to the GFDL, I think in my mind, that's an example of a disagreement that will not in the end be a blocker to endorsement of Debian by the FSF or vice versa. I think that we understand each other's perspective on that. And it's not something that needs to be an obstacle to lots of other forms of cooperation. And I think that we get on the same page about more things and then we come back to that and see what we can agree on. Thank you. Other questions? It's time to wake up. Well, this is not really related to Debian, so I was waiting to see if other people had more other questions. But regarding the director that you have for software, for distributions, the endorsement for distributions, do you have anything for services? Like, that users, even if they have a free system, but they agree in terms of services and the cloud. Yeah. So how the users can believe on FSF or any other director or organization that will support us in making a decision. Yeah, I think that's a very important issue and I had had it on the, you know, our approach primarily is to try to encourage development of decentralized replacements, decentralized free software replacements for the services that people have been using. So I think that that's something that we can work together on also. You know, a big problem with why people don't already use the ones that are out there like Media Goblin or GNU Social or Pump I O is that they are not easy enough to set up. And obviously as a distribution, you know, Debian has a big role to play in helping make that possible. So that's one answer is to try to make the world better for users that don't want to use the centralized services. And then within the directory, we have a way of creating different kinds of collections. So we can tie those programs together as these are programs that are related to the problem of network services or software as a service. And then we soon will be publishing some recommendations and criteria for specifically hosting services, what kinds of criteria they should meet in order to be used by GNU projects or recommended by GNU or the FSF. That's kind of one category of service that was really in the news after Gatorius went away and people were sort of like, where do we go now if we don't want to go to a more centralized place like GitHub? So those are kind of currently what we have in mind, but definitely there's a need for talking about user freedom outside of the programs which run on their individual computer and how they experience that on a network. Thank you. You previously talked a bit about hardware supporting free software and free software supporting hardware. I wanted to know what does the FSF thinks about the participation of Steam and Debian and how even though Steam uses proprietary hardware, proprietary software, how can this help bringing free software to hardware? Yeah, I mean obviously we don't like Steam OS as a distribution because it does contain proprietary software. It's possible that in the end it may result in games or a problem for free software users. I can understand the perspective that it is an improvement over having a dual boot windows environment in order to play games. And I know that Steam has been and Valve have been very cooperative, technically speaking with the Debian community. I think that we should continue to ask them to release the Steam code itself as free software and also to make sure there's room within any marketplace of software that people are using for free applications which I'm not sure that that's possible in Steam right now. So I think we definitely need more initiatives around free games. We do some small things at the FSF. We run a LibrePlanet gaming server that you can get involved in administering or playing games on and it has mostly some of the classic multiplayer free software games. But it's a topic we're interested in. It's just we need to find ways to compete with a lot of social aspects like the fact that video game companies get subsidies to make proprietary software and the industry is extremely competitive that way. So there's a lot of problems to tackle there. Steam may end up in the end being helpful but I think in the short term it's not something that we can recommend users to install. Yeah I just wanted to kind of expand a bit on what we said a moment ago about well essentially Cloudy stuff, software as a service, things like that. This is something that Debian is coming to recognize I think as one of the biggest threats to user freedom in general. Most of the people here in the room will probably have a non free phone running all sorts of interesting applications and it's quite difficult at the moment to do better with that. I know the FSF are slightly going in that direction. There's some projects within the FSF but I think I'm slightly concerned at your list of strategic goals doesn't seem to have that as a thing that really needs to be addressed and you want to be looking at that much more closely. Yeah I think that in the discussion it is implied by a couple of the goals one of which like you said that there's a very tight connection between the mobile environment and the service environment. These services are getting a lot of traction because on smaller devices you wanna be able, well you have multiple devices so you wanna be able to access your files from multiple different places. You want to be able to participate in services like Uber, airline applications and all these things that are geared toward life in a mobile sense. So I think addressing the mobile side of things having all those applications be free software is a precondition to effectively addressing the service question. And then I think that's the issue of having the general public see free software as their issue should probably, I should have actually written that as the general public to see user freedom as their issue. And I think that that is getting people to care about what happens with their, when they give up control over the software that they would in the past have run on their system to a remote server that whether that software on the server is free or not doesn't matter because the user using it doesn't have access to the code or to modify the code or do anything with it. So I feel like the goal of getting people to see their freedom as an issue strongly implies figuring out how to replace that remote service model. But probably it should be set explicitly there, yeah. Thank you. I recall from earlier bits from the DPR letters that there was some ongoing discussion between you and the DPR about solving the endorsement problem. Can you give a status of, or maybe can you give your view of the effectiveness of these discussions and whether they are stored or they are still ongoing? So yes, so that started when Zach was DPL and has continued since then. And I think that they have been, I mean, they're why I'm here talking about this. You know, I think that those conversations made me feel very welcome to come and have a conversation and to talk with people in addition to DPL and the Debian community about these things. So I think they've been very, from my side, they've been very effective at getting things started back up again. They have been slow, but I will accept basically all of the blame for that. I have not in any case been waiting for a reply from the Debian side. So they have been slow, but that's my fault. And anybody who's been on that mailing list for quite a while knows that there was kind of a burst of activity at the beginning and then I didn't follow up on what I was supposed to do and it kind of trailed off. So what I've been trying to do by coming last year and this year is to commit myself again to helping those conversations move forward. But I really appreciate the support that I've gotten from the Debian side and all the DPLs. Thank you. Thank you very much, John.