 What? You use Facebook? Lots of people use Facebook. I've been walking around the aisles, watching people use their Facebook while they're listening to the lectures. So, but even the, clearly even for people who don't use Facebook, I think there's a market for the Freedom Box, right? The Freedom Box. You clearly missed the big lecture this morning. I mean, Evan, Ben Moglund's talk this morning was, I thought it was pretty interesting. I'm curious. I guess there's gonna be a BOF about that that's getting scheduled right now that I'm supposed to make an announcement about later on tonight and I guess there's, I mean, who's, show of hands, who is interested in, okay, two questions. Who's interested in owning such a thing? All right, that's pretty good. Who's interested in developing for such a thing? That's cool. Yeah, okay. For those who don't know, basically the whole focus of the talk this morning was a proposal for this device called the Freedom Box, which is essentially a server in a small embedded device that you plug in in your home and it serves all of your whatever information you want to be public in the world. So, for instance, your home page, your, you know, free Facebook equivalent social networking sort of thing. It serves as a network proxy, telephony, which was actually what this talk was supposed to be about, and privately owned router, right. So something that would be running completely free software on a embedded device that you would plug in in your home and it would own all of your data, would control all of your data and how that data is distributed on the network, and then also act potentially, I mean, in Professor Moglen's dream as, you know, a proxy for other people and, you know, Tor Router or whatever. Consumer friendly, dead simple. And his, I thought the talk was interesting, was very interesting because he had such a concrete proposal about what to do, and the point was that it, you know, would exist on, you know, hardware that currently exists right now and, you know, with built, operating system built with stuff that we already have, like Debian. So, well, it would be, well, it would be dead simple to the, the idea is that it would be dead simple to the user, but it needs, you know, however many hundreds of smart, well, the idea is that it wouldn't need a system administrator. That's the hope. That's the ideal. It's just that a box that you would just plug into the wall, and it would just work. I mean, clearly that needs a lot of work for that to happen because that sort of thing just doesn't even exist. All the things that claim to do that clearly don't do that, like your, you know, your home, well, I think the thing that's closest to that is your home router, your home gateway, without wi-fi. Those things generally just work, I would say, but as soon as you add wi-fi, then they, they stop working. Yeah, here's the microphone. Has Mr. Sugar showed up yet? No, okay. I'm gonna keep ranting on this, then. Does anyone know if the chipset in those, those boxes? I think he was, by the way, I think the hardware that he was referring to is probably the guru plug. That's what it sounded like. Yeah, I think that's right. Does anyone know if that chipset supports host AP yet, though? Supports what? So it can be an access point? Because a lot of the chipset still don't support that, right? Those things are access points out of the box, I believe. Oh, okay. Right? Somebody, there are people who own guru plugs around here, aren't they? They have bluetooth and wireless? Hello. Yeah, I work with these devices and on the arm side. Yeah. And there's the guru plug, the first one didn't have wireless. And now they're shipping, they call, it was the Shiva plug without wireless. Then there's the guru plug, it has wireless. These devices are based on ARM V5 instruction set. It doesn't do much graphics. And there are a lot of devices coming up, like EFICA MX boards. Also, they are based on ARM V7 instruction sets. They can do multimedia and have Wi-Fi. And you have all the software from Debian to make an access point to have your server at home or do whatever you like. I mean, I've been running a system which is used for Debian building these at home. And I have my server there with my IRC connected, have my website, my documents. And I can use it as I want. And this is not, for me it's not new. You could do this for like three, four years. And now there's more, there's this plug computing and there's like a hype on these devices. And I mean, I think this is great. Yeah, I think that probably most Debian people, I mean, I'm sure everybody, most people in the room like maintain servers in their room. So I mean, he actually mentioned that, that most of us probably already do basically what he was talking about ourselves. But the idea is to make it something that's dead simple that anybody could just use and plug in. And so clearly the challenge there is not the, since I'm being the speaker now. It's clearly, clearly since the idea is, I mean, clearly the challenge is to make the actual thing that's going to run on the device, not the device itself, since that's done, obviously. So yeah, go ahead. I just met a gentleman next door. He's working on a project called Linaro, it showed it to me. And what they want to do is they want to coalesce all development efforts regarding the ARM platform, providing APIs, everything so that everything from Mimeo and Android, everything can be found there in one place, making coalescing all APIs and making development on the ARM very easy, in other words. So that could be something that could be used with this project to move forward towards the project that he was talking about this morning. Yeah, I think that, I mean it sounds to me like what he's really asking us to do is to make a Debian, essentially a spin-off distribution where you make a installable thing that you just can load into one of these devices and it just works. Something like Ubuntu where it's, you get a core, not just a core set of packages, but also all the way up through the very top such that you've got all of the components already there, already installed when you put the image onto the device. It doesn't have to be a complete spin-off. You just have to have enough meta-packages and a way of putting those together as a bootstrap environment. So you can do it with the multi-trap that we've been looking at already for this kind of thing. You just select the packages you want and you put them together, you put, you give it its own configuration and you have a turbo at the end. We even have a name for that. It's called Debian pure blends. Well actually Neil, you're actually the, seem like the person of the content. I mean you're the one who's been organizing the Mdebian stuff, right? So I mean, is that, how close is that to being able to provide, well how close is that being able to provide a system that'll just run? It is in commercial production today. Oh okay, well there you go. But clearly we still need the integrated experience of having the pre-configured services and stuff that would run in the system. Yeah, go ahead. About six years ago I bought a machine called a Redi NAS and that company was bought by Netgear and it came pre-installed with the version of Debian for like home file storage and it was pretty much, it was pretty much what we're talking about. You know, you bought a machine, it was pre-installed with Debian, they gave you the root password if you wanted to install anything else. I didn't think it was an Intel architecture either. I think it was something strange. Okay, oh yeah, Spark, it's a Spark architecture and I think that the way that they had that set up, you know, was like a proven business model and that there's probably room for lots of small companies all around the world to do the similar thing. And these, the kind of Shiva plug computers are very low, you know, with ARM, are pretty powerful for the given that they're the size of a power supply. But I think that yeah, you know, lots of people have got old computers sticking around that they could use for doing this kind of thing. And so I think the idea of it being one specific like Shiva plug thing isn't that relevant. I mean if the packages are in Debian. Yeah, no, I think that the particular, I agree with you. I think the particular hardware implementation is not as important as the software stack, the operating system or whatever you want to call it that gets installed on the thing and having that be a unified top to bottom thing that works. Yeah, and installing these things is something that small companies can copy the ready NAS model and do likely. So the next step seems to be to me kind of figuring out what stuff, what is in that stack, right? So clearly, well, do you guys know about the diaspora project? He mentioned it. I mean, you might have, anybody, raise your hand if you heard of this. So this project, I'll give a very brief overview of what happened. So this project actually spun off of the lecture that he gave on this topic in February. And it just for NYU kids decided they were going to basically try to implement the social networking component of what he was talking about. And they started a Kickstarter account and the Kickstarter account drew in like $200,000 and it blew away by like an order of magnitude everything that all the other projects that Kickstarter had ever hosted before. And so now these four guys are often in San Francisco basically trying to implement this thing. But they're trying to implement essentially the social networking component of that, which would be something that, you know, I don't know, they're still trying to work it out. But basically a service that runs on this embedded device that you would maintain control over or ideally and then, you know, gathers information from particular other nodes in a distributed manner that you have identified as nodes that you want to have communication with and who you want to share data with. Excuse me. I think now would be about SIP and asterisk, right? Yeah. This is super cool and super interesting, but I'm just thinking maybe we should like, are we all agreeing that this is, that we have a high tech... I don't know. I was just doing it since the speaker never showed up. The speaker won't show up because he hasn't arrived at DevCon. So he won't show up. So how about this? One of the things that he was proposing talking about being on this device was a SIP asterisk server. So clearly, hopefully people are in here at the room are interested in that topic. Yeah. So maybe I can jump in here. I missed the beginning of the discussion. So one of the things we wanted, we had an informal buff after the session this morning, the presentation by Morgan, where we tried to coordinate a little work team to work on that and it by no means was meant to be restrictive or whatever. But we now have a wiki page in the Wiki. It's called Freedom Box. So people want to look it up. It's there. We wanted to organize a buff for this, but it seems that it's self-organized already. So that's fine. So we can like talk about all this. But one of the things we have already done is tried to come up with a mission statement or a vision statement. So if people want to look it up right now, it already has some good grounds. I can't read it because I don't have it on my computer. But my point is that the idea behind all this is to really create an integrated stack of software with a common, very easy to use interface that will allow to do multiple things, including VoIP, including email, including software application services at home, and all this stuff. So for me, it's a very important project that I've already, that I was already interested in. And for me, what even has changed is that he's officially promoting it and getting people interested in the project, which means that a lot of people are going to stop, well, not stop, but instead of just doing it all on their own in their basement, which I've been doing for the last 10 years, but they are going to stop sharing that expertise and creating a distribution that anybody can install on whatever machine, like the Shiva plug is a good example because people can relate to it more easily than installing a crappy old gray 486 or Pentium or whatever it is now that is a crappy computer. So people can more easily relate to that. But basically, the idea is to create what we said earlier, like a custom demo distribution or a bunch of meta packages that are going to bundle everything together for everyone to make it easy to install, easy to set up, easy to use. Hello. Yes. Did you want to describe, this is, I guess, an example of the form factor. This seems to have an actual LCD screen on it. What is this exactly? Does somebody who knows what this is want to describe what it is? This is a balloon board. Wait, let's get you a microphone. This is a balloon board. It's attached. Well, this is a development platform. It's an ARM device. It has a PXA board. It's open, so you can show the board. Does this good for people? Can people look at this? Can I pass it around? Yeah, you can pass it around. Be careful because the board is loose. It's a little bit not flaky, but there's one in a box. And then it has an LCD. And this is a development platform we use. And it could be like the freedom box or something like that. It's willing to show. And there's some other devices. People have like efficacy mix or maybe you want to show. The only catch is it's actually quite hard to buy a balloon board. You can ask us nicely and we'll keep not getting around to sending you one. How similar is this to the, oh, is that, is this a similar device as well? Oh, okay. So that's an, that's an EffiCat MX, which is a rather newer board. It's Cortex A8 base. So that's a modern ARM V7 current generation. How closely related are these to the Guru plug? The Guru plug is an ARM V6. So that's about in between halfway. But it's the Guru plugs quite a lot faster. That's a gigahertz, I think. So it's got, well, so, you know, the EffiCat MX has got quite a lot more welly than a balloon board has, but balloon boards got more memory and flash on board. You know, you pay your money and takes your choice, right? There's an awful lot of these boards. Fundamentally, as you say, the hardware doesn't actually matter that much. Right. I think it's interesting that there's an LCD on that, on the balloon board device though. Is that just? You can plug it into a TV version or? We've done a TV adapter board as well. So you can actually plug a good old fashioned monitor in the back if you really want to. The problem with that is that it uses almost all the memory bandwidth that's plating your screen. So it's not really very useful. So getting back to the SIP telephony asterisk issue. Is there another question? Yes. Go for it, Vinner. Hang on a second. There you are. Nicely. Don't move now. Yeah. Well, yes. I went also to go back to the talk topic and, well, in any way, still related to this presentation in the morning. I think for most of us, well, we don't really need much of a Skype replacement because we have several. I mean, I use a key on a frequent basis and I know that although I haven't successfully used it, I know that Pigeon already supports voice as well. But the problem here again is the end users. Yes. And not so much in terms of a hard to use interface, but as for network friendly protocols, which is quite a problem with SIP even more. I often do video conferencing at my university. H323 really, really, really sucks when used from non-public IPs or so. And, well, I have been trying to get some friendlier programs, standards, protocols that my non-computer literate, well, let's say clients can use to connect to participate not even in a one-to-one meeting, but a one-to-many, which is quite problematic with the applications we now have. We have something called OpenMCU, which is an audio and video call multiplexer, which is packaged in Debian. But as far as I could test it, it works fine when one person connects. It still works mostly when two people connect, but then a third person connects and it dies. So, I mean, clearly, it's no improvement overall. So, I'm curious about what's the interface meant to be for the phone part of this? Well, yeah, yeah. The problem is not the interface. The problem is that most home users or most small enterprises will not have a full-blown public IP. At least in my country, in many places, I've seen you have non-rotable networks mass-created. Or, yeah, but you can control it. Yeah, IPv6 would be good, but I cannot control what the provider gives the customers. So, one of the problems we have right now with SIP, and even I don't know if any of you here run a SIP server at home. One, people, two, a few, okay. So, you probably experimented with the various problems related to NAT and firewalls and all this stuff. SIP isn't denying to deal with heterogeneous networks and stuff like that. So, one of the things I think, one of the objectives of the SIP, which project, is to work around those issues, the routing problems we have on the Internet due to censorship, due to restrictive ISPs that break our networks and things like that. From my understanding, the other objective of the project is also to provide cryptography and security and confidentiality to the calls. Because SIP, as it is right now, is completely in the clear, has no security whatsoever. And there are extensions to try to fix that, but it's always layers of crap over existing crap, basically. So, it's always something that's much harder to adopt than if you start from scratch and say, okay, we're gonna replace Skype. And I agree that I personally don't use Skype, but I would really appreciate having that decentralized network available for me to work my way around restrictions. Right now, if I wanted to create it, do a SIP connection here, I don't know if I'll be able to, because there's maybe a NAT, maybe there's some policies here that make it impossible to do that. So, I think there is a requirement, there is a need for such a software to be available. And I think in general, one of the ideas behind the Freedom Box is to have all the services we have decentralized and secure, so that we bring back the original ideas of having a secure network and a decentralized network at all parts of the infrastructure. Not only, like, well, I was about saying, not only email, but even email is not secure, right? Because we don't encrypt our emails, we rely on, like, the crappy security email provides. So, if we try to provide tools for people that bundle all those secure alternatives, then people are going to start using them because, not because they're worried about their privacy, because obviously they don't worry enough to change their habits. But because it's cool, or it's the latest thing, or it's, you know, bringing back, like, a better tool to, as I was saying, replace your wireless router with this little cool gizmo with a monkey on it, or whatever. Because they've got LCDs now, which the guru plug doesn't have, and I think the LCD can go a long way. And I think you have a very valid point about the interface. Like, a phone is a very simple user interface, and when it got in, when it started, people started using it, some people weren't able to use it because they couldn't remember the phone numbers, they couldn't know how to type it in there. It's not necessarily intuitive interface. We consider it intuitive because we're born with it and we've been taught how to use it. But a phone already isn't quite easy. Yeah, I mean, you're talking about, you know, if people could remember the IPv6 addresses of their friends or something like that, then it wouldn't type it in. So when we talk about SIP, we talk about email addresses, we talk about something already much more complicated than phones. And if we were talking about an interface, then it means that we need to have either a phone device that you pulled here, here or something. But you still need to figure out how to dial the person you're going to dial. Exactly. That requires some screen or something. So one of the things you were talking about was to have a cell phone connecting to your device, so that instead of relying on the device itself doing everything, the device is kind of a focal point. It's a server that you have multiple clients attached to it. Yeah, when you're away from your home, you would connect to your home to get back to the other people. Yeah, I think to follow up on that, that's, I think, highlights a key point about the freedom box idea is that it's not going to be your phone, it's going to be your phone server. And things like Android phones, like Memo, I mean, whatever it's called now, Amigo, phones have Voight built in. I mean, there is still the problem of what you're talking about that SIP kind of sucks over the internet, and Skype has done a really good job of like penetrating all that. But there was a lot of discussion about the interface, and like this thing is kind of in the middle. It's like your email box. It's your phone server. It's your Facebook server. It's not the interface to your Facebook. It's not the interface to your phone. It's not the interface to your mail. That you can use other things. So one of the things that worries me about the situation is that most of these services that we rely on right now, we work because of high availability. We rely on email services and phone services that are essentially always on. But I don't know about you guys, but my home network is flaky as all hell. And it's supposed to be, from Time Warner, it's not a sophisticated business thing. But in order for this device to work as advertised, you need a good reliable net connection. So that's going to be, that's I think going to be the hardest thing to overcome since we're not going to be in control of that at all. And Evan was, you know, he was pretty adamant that sort of if we build it, they will come sort of attitude. And if we build it, we can push them to do things, them being all the other things that we don't control. But you know, there's some pretty big challenges there, especially since we're one of the big people that we're going to be fighting against is our upstream providers who are not going to be necessarily happy about this or into it, you know, in whatever form it takes. But like the point, there was a question raised about, you know, ports that ISPs usually block a whole bunch of stuff that they don't want people to stuff from. So I mean, when I first moved to New York, I had an unbelievable battle with Verizon because flat out denied that they were blocking port 80 from my apartment. And I was like, you guys are a pack of liars. And they just said, no, we're not doing it. We're not doing it. You're fucking lying to me. I can't believe you're lying to me. And I finally got to like some upper manager after like three days and said, oh, yeah, we're block. Fuck you. You just like, I just spent three days on the phone battling you about this. And they wouldn't even admit that they were doing it. So I don't I don't know how we're going to necessarily overcome that sort of thing because that stuff's going to have to be dealt with for any of this to work as he's outlined it. I just wanted to go back to the comment regarding remembering your buddy's IPv6 address. Because I think for any technology to really be successful, it's the reason that cell phones have taken over on smartphones and things like that is because they're extremely convenient that they they basically decrease cognitive dissonance. And I think we may find ourselves getting into, you know, it FRC's and stuff because what it seems like what you really want is you want an ID. I know this guy. He's team answer with that mean dot org. And there's a DNS server that that has a lot of things on it, including something encrypted web finger. That's just me. I just thought of that. Okay. You guys heard about this web finger thing that people are talking about. Okay. So is does it include some way where I can encrypt with all of my friends? Public keys. Thank you. So that because what I want to do, I want to put a lot of information about myself out there, but I only want folks whom I know to be able to get to it and not including Zuckerberg from in fact, actually, some people that I work with have been talking to the diaspora folks about that exact issue actually how to use public key cryptography to make sure that those those like point to point connections are secure and authenticated and that sort of thing. I have I have another thing to throw into this whole puzzle, which is when we're talking zip, as I guess some of us here is interested in is the latency problem. Because one of the things that I fear that upstreams will fight us with is to optimize even more aggressively for bandwidth, which means that it kills the latency so that we cannot do interactive. What I what I what we have today in Denmark is that the biggest provider is providing this little five port Ethernet box with direct ADS DSL inside it. So you have to do their box. You cannot open it. You don't know the password for that ESL encryption and this box, one of the parts if you pay for it has optimizing for a latency so that you can use it for what and the other ones do not and things like that. So you have provided their phone connection. It's a phone provider. They also sell DSL lines, but then they deliberately lower the latency of the other parts. So what I could imagine that is easily done for them is to just make it clutter so that you always get a delay long enough that you don't want to use it for sip. Yeah, especially if they're trying to sell you phone service. Exactly. So I wanted to comment quickly on two things. The first one is the question of identity. Basically, what you were raising like touched on earlier for me, phone numbers are going down the drain in my mind. Email addresses are way easier to remember and have actual meaning and work already. They're used for sip. They're used for email. You can use the domain part for a website like it makes sense. So that can already be used. So for me, that problem is technically resolved. It's not a technical problem. It's a social. It's the question of who owns the domain names, managers, all this stuff. I feel like there's a little bit of a technical challenge there, but that who knows about webfinger? Who's heard about webfinger? Oh, wow, surprisingly few people. I thought I was a late comer to this. Both? Oh, it's the same thing. I don't know. Who knows about both? Not that many more. So briefly, what the idea with webfinger is from what I understand is that basically, email addresses become resolvable. Like, they become something like a DNS record where you can, using somebody's email address, you can find out information about the sort of services that person provides on the net. Maybe somebody who knows about it more can correct me if I'm wrong. I'll add a small thing. The way I understand it, webfinger is a slightly distortion of a different approach to the same as Fof plus SSL, which is the inventor of the web who is now working on the extension of using the HTTP layer itself. So it's not email addresses, it's web address, it's URLs. So it's a web address which then you tag as also being your identity. Right, but you have to somehow resolve the email address to a number or to a URL. Yes, and that is the reason I mentioned Fof's friend of a friend network. So the networking relationship and the linking into email addresses and other resources on the internet is done through the RDF databases. Well, it sort of sounds like a panacea to a lot of these issues about identifying where your friends are on the web. As long as you know somebody's email address and you can webfinger that email address and figure out where their freedom box is and that sort of thing, then it makes that a lot easier. So if you want to make phone calls or whatever. Is there a microphone on this side? Oh yeah, go ahead. And reference to what you mentioned before in terms of Verizon, fighting with Verizon. Verizon is hell. So basically, unless you're going to go through an ISP that's going to deal with Verizon for you, which I've done before, you're probably going to have to go to DSL reports and do a search for a provider that's going to give your service. Cable service sucks, but if you want to have your ports open, you probably would have to go with DSL. That used to be speakeasy, but I've been pointed out. There are lots of different companies. I've never used speakeasy. I use, for a long time, I use Cyberronic, they're located in Boston. They're DHCP service. You can get 6 down and 768 up. So you can get higher service, but it's going to be a little more costly. Okay. Well, yeah, he was speaking about Verizon. And that's the thing that came to me. Yeah, it's an example. And so basically, they don't block your ports. You have less chances of having these independent companies not blocking your port than these large corporations in terms of Quest, Verizon, Comcast, have a tendency to want to block ports because they're also irresponsible people running applications, things that they're trying or misconfiguring web servers or especially mail servers. That's why they usually are blocking out ports. What I wanted to say to him is that if you want, if we want to circumvent, I'm going to speak about the US because that's what I'm most familiar with. If you want to circumvent certain vendors like Verizon, you might want to go, you might want to look for smaller companies, which are, which are going to give you the service between $40 and $100. You're going to pay more in cable, but you're going to be able to use all your ports and do everything. There's no limitation as far as I've used with them. And I've had them for years. No limitation whatsoever. You can run any server you want. That's true. But it also makes it the adoption issue is harder. If you want to get adoption, it's harder to say, oh, buy this box and find a service provider that does this. Here's a list, but oh, and you're going to pay twice as much. And those are all difficult. Because of the core list, the coalescing of services with these large companies that are buying up smaller companies. And so you have Time Warner. Time Warner doesn't allow cable vision, for example, to get into their zone. So there's no need for competition. They don't need to improve service at all. Verizon has their own stuff. And then it becomes very difficult. Basically, those are social political problems. My first point was that the identity problem was not really a technical problem. We have a lot of solutions. It's mostly a matter of adoption. And I think that the fight for network neutrality, which is what we're talking about here, goes beyond just which provider you choose or which board is blocked or how they walk with your latency and all this stuff. As long as we're just a bunch of geeks that want to play with asterisk at home, they're not going to bother. Nothing's going to change. When we have the freedom box, when we start operating those systems, then we have leverage. Then we have leverage to start asking for things saying, people want to own their own infrastructure. People want to operate their own services at home, and this is legitimate. This is not just a bunch of geeks in their basement. Everybody uses them. Everybody tries that. And some providers do it, but they're getting rammed down by Verizon or by Durge Telecom or by whatever company out there. So the idea is to have leverage. The idea is to have a good use case on how the internet should be. Because a lot of people that make those policy decisions figured that the internet is like a television or a series of tube or whatever, that you just sit in front of and you get fed stuff. That's a lot of the vision that's in public administration and that's partly true to some extent. Like the way Amazon or Google or a lot of companies operate on the network is that they build those gigantic websites and those huge data centers and they own the data and they push it to you. And you give them a little, well a lot of your personal information, but then they own it and then they push it back to you. It's the same way that happened with television or radio back then is that companies own the network and they own the data and they own the information. And by changing that, then we start talking about the rules and why we want to change the rules. And that's why this whole thing is interesting. I think that as long as we just say, oh, they're blocking my ports, that's too bad or I'm going to change providers, some countries in the world right now have 100 megabit at home and don't have those kinds of problems. We need to demand better services. This is ridiculous right now. We're completely capable of offering proper internet connectivity at home and we're not doing it because those companies are in conflict of interest. There used to be laws against that, against monopolies. UD, AT&T was broken up back then and it's reunifying everything. So those things are policy decisions that are made at a higher level in every country in the world right now and those things are things that can change and having proper leverage for that will help. Going back to the identity issue, it seemed to me, I was thinking about this earlier, as several of us have been. If you've got a dynamic IP, which I imagine most people have, you've got to have some way of being able to get that information out onto the net. I don't know if this both thing helps, I don't know anything about it, but it seems to me that it's impossible not to rely on some external service in order to broadcast your IP address at least so that you can get back to the box to say this is the box responsible for this particular identity and therefore the authentication of whatever services are provided. And then the question arises, what is that service? We're going to use dindianes.org for this, they're tiresome people. Clearly we need IPv6 now. Does that solve that problem? I think so. Well, you don't need because you don't need to have a dynamic IP address with IPv6. If your IP address is not dynamic, then it's not an issue. You just publish the mapping from your email address or your URL to your IP address and then people know where to find you, right? That's great if it works, but I suspect we're going to have a transition period where in practice we haven't got that and we're going to have to, okay, I guess if it's any transition, that's cool, we'll just have to put up with some intermediary for a time. Well, Debian is pushing IPv6 at least as well as anybody else's, maybe better. So obviously we're in a position to push that forward more. I wanted to kind of bring it back to, I guess, what's on the subject of this talk, which is replacing Skype. And Skype actually is actually for more of the ISP discussion that we just had is a really good model. As far as I know, no ISPs have kicked people out for using Skype. Skype is a peer-to-peer service. It is serving. Skype is like many network administrators hate it because it looks for any open port and it just, it doesn't obey the rules of, you know, this port does this service. It just is like, I'm going to do whatever I can to get a working phone call. So we can do that as well. And when you look at ISPs have contracts, so this is something I think it applies in most the world. I don't know about it, but the contracts say you can't do this. So if you violate the contract, they terminate it. Maybe they find you. You don't get arrested. I don't think there's many places in the world right now yet, but I don't think there's many places in the world where you'll be arrested or really much punished for violating the terms of your ISP, terms of service. So I don't, I don't feel like that's the worry though, right? I mean, just having your service terminated is bad enough, I think. If you have no other options, yes. I mean, if plugging in this device got your service terminated in a couple of days because it was doing something against your ISP terminated service. Right, so can we go back to Skype? Does any ISP terminated service because of Skype? Skype abuses, Skype is serving. Skype is abusing a lot of rules. Yeah. Okay, most of the world, I don't think so. Yes, I don't know if who here has actually studied how it is at Skype works. It's not quite correct to say that Skype is peer-to-peer. It can be peer-to-peer. Most of the time, it's peer-to-proxy. And so I'm leery of any suggestion that relies upon some external thing like, you know, when the Great Pumpkin comes and IPv6 is available. So, you know, how long have we been hearing about that one? So, I mean, it seems like the key elements are like a distributed hash table with supersedes, which is the way that Skype works, is that there's a set of known hosts that are guaranteed to be on and a community could provide those. And then, you know, you basically, when your box logs on, if you have something that's not a nasty, you know, blocked port or something and you basically, you just sort of set yourself up as something like a secondary, and then it just goes from there. Either you try to find out the address from a super or one of the cached secondaries and then off you go. So, it seems like that in combination with some of the work that Dan Bornstein has been doing with high-speed cryptography, I don't know if anyone has looked at DNS curve at all. It looks like some very interesting stuff. So, I think a hybrid of some of what Dan has been doing with the basic idea of the distributed hash table and the peer-to-peer, if possible, but typical case peer-to-proxy could actually be quite viable. Apparently, we only have 10 more minutes left of this lecture. Hello. Yes. Yeah, there's a question. Well, we've been debating on IRC as well, and there's a question from Timo Lindford, Timo Lindford from Finland. And do you really think IP version 6 is going to help with NAT firewall problems? And if it does, what about mobile IP, moving around and using wireless LANs with your PDA and going through? I certainly can't answer that question. It's not a question for you. It's just for the audience. Of course. And also, you have to go through all these firewalls and all these things. It may be useful to design some kind of library or something that can get through all these... Well, certainly the point of all in the network is that we don't have NAT anymore. That's sort of like baseline point of IPv6 is that we don't have to deal with NAT. But firewalls, I think, you still have to deal with. Go ahead. Is the light on at the bottom? Hold down the button. I just wanted to reiterate something that was said this morning, and that is the target audience of this project is really not geeks, because many of us in this room can already do this. And so we need to make sure that this is familiar and usable by people on average ISPs with average connections. And it's only by being usable with what people have now that will gain the leverage to be able to change things. So we have to make sure that, for instance, if the average person wants to use a device to be able to make calls to people that don't have VoIP or don't have a freedom box, that we can provide some sort of interface to do that, maybe through any number of SIP providers out there or whatever else if SIP doesn't work. But similarly, we need to make it easy and with something that they can understand right now. One example would be you can get an analog telephone adapter. They're less than $100. You can plug a regular, any old-fashioned phone into it. There's your interface. You know, more expensive than a soft phone. Except for the dialpad part, right? That's what I was worried about before. I mean, if you don't know where your friends, if you don't know what their friends IPv6 address is, or whatever their internet address is, it's hard to dial it in. You need to have something to Right. If you're going to call somebody. Yeah, you need some mapping. If you're going to call over the internet some way to let them punch in a phone number or something. But I think that a lot of people are going to want to not just talk to other people with a freedom box. And so, what can we do to make this exciting for that kind of person too? And can we let these third-party vendors that do sip termination get involved and have a marketplace or something out there? So, I wanted to answer the RSC question. I think regarding IPv6, it's just one of the pieces in the puzzle. I don't think it's going to solve all the problems. People always make fun of IPv6 as coming real soon now. But actually, IPv4 is going away now. That's what's happening. We're running out of IPs. This is going to happen in one or two years now. It's a very clear and prison danger, and it's going to change a lot of the things that are possible right now. Right now, you can just get a server, get an IP. It's cheap. An IP is cheap right now. You get an IP at home. You get usually a real IP at home. You're usually not netted, at least in North America, as far as I know. There is something that do it now. So, that's changed already from my perspective. So, those things are going to change. So, we are going to require some change in infrastructure. And the reason why we talk about IPv6 is not because we're geeks that want to play around with little toys. It's because we're technologists. We're scientists. We're people that actually build the infrastructure. We're the people that are building the future technology. And if we keep on doing the right thing, then the right thing may happen. We need to stop whipping ourselves and saying, oh, we suck, or this is never going to happen. We need to start building those things. And I think it's cheesy to say that, but we need to just do it. And I think if we build it, yes, they will come. And I think the Freedom Box is exactly that. It's one of those tools that we're going to build on. The thing that was mentioned about mobile IPs is, for me, is missing the point. Of course, phones are becoming more and more powerful and they're becoming general purpose computers. But they're always computers. You don't host a server on your laptop. You can, but it's just a tool. I don't host my personal emails on my laptop because it's moving all the time. It's not convenient. And if I lose my laptop, then I lose my emails. So my emails are at home. So my IP is at home. It may change. That doesn't matter. I used to be on the dynamic IP for a long time with the DNS. It was working fine. I changed just because it was more convenient today. It's actually a more security problem to have a static IP. So I'm thinking of changing back, but my IP is kind of cute, so I keep it. But basically, this is irrelevant. This is all very simple technological problem we need to solve. And what we need is to have the proper leverage, the proper tools to offer people so they can start playing with new tools, new toys that anybody can use and to have those cute things that people want. The new Gmail, like Gmail at home. It's better than Gmail. It's your email. You can offer a Gmail to a friend. You can offer a phone number to a friend. You can provide a gateway to the phone system to your friends because you have a phone line at home and you hook up the Shiva plug to it and then boom, you're a gateway to the traditional phone system until it dies. But this is all stuff that's doable right now. This is stuff I've been doing for a while. And this is stuff that everybody here has been playing with. It's just that we need to start sharing it instead of like acting like basement geeks and like, I've got the boy working now. We need to share this stuff now. We need to share this stuff and make it accessible to everybody. That's the real concern we have right now. And if we stay in our basement, then we're going to stay in our basement. We need to get out. Yeah, I agree. You're here. So I think that's probably as good a place to stop as any