 So thank you all for coming to hear me talk today. I'm going to talk a little bit about the GPL. Specifically, I'm going to talk about stuff that we don't generally talk about the GPL. We're late enough in the GPL process. It's done. The license is out the door. And we've given up on being the we're no longer the people who are involved in the crafting of the license. And we're not even the people who are involved in advocating for the license, because that's what the Free Software Foundation does. They were our client during the process. And having finished the job of helping them write the license, we don't really have a huge role in forming the GPL at this point. And so when we talk about the GPL, we don't necessarily talk about it in an advocacy context. But what I want to talk about today is more about the things that interest me about the GPL and what interests me very much about the GPL at this point is what we failed to accomplish in the GPL process and the limitations of what the GPL has done, can do, and where that leaves us as a free software community. Usually, when I talk about the GPL, I give people a choice. Do you want to hear the rah-rah? I love the GPL speech. Or do you want to hear the, well, I have some reservations speech. And not once has anyone asked me to cheerlead for the GPL. So I've stopped asking people if they want to hear the cheerleading section. But if there's anyone who has any doubts about how great the GPL is and they want to hear the cheerleading speech, you're welcome to come talk to me afterwards, because I never tire of telling people how great the license is. So first of all, what is the Software Freedom Law Center? I just want to give you the quick background on that. We are a nonprofit pro bono law firm. We're a law firm in that our work is client driven. It's all client service. We're not an advocacy organization with an axe to grind so much as we are a support organization for people that produce free software. And we're pro bono in that we don't charge our clients for the work that we do. And we do it all for the public good. And we're a nonprofit law firm, which means that we get to starve with dignity. And we enjoy that role. It lets us work on really cool projects like the GPL. So we, as a firm, spent what we initially had envisioned would be about a year of public debate while crafting a license with a lot of input. And it stretched a little bit longer than that. But on the whole, we regard the project as quite a success. We produced a really good license. But what is it that we produced? We produced a copyright license. The GPL has always been a copyright license. People write software. They have copyrights in that software. And they want to give other people permissions to use that software. And those permissions are conditional. And the conditions are laid out in the GPL. The GPL is also a patent license, which is a new feature of the GPL. GPL version three has an explicit patent grant. The prior version had an implied patent grant. And this is a new step that we thought was a good step. But it also is a complicating step. And I'll get back to that a little bit later. So we've got a license that is a patent license. And it is a copyright license. It is not a trademark license. And that also causes some problems, which we'll talk about a little bit later. So what isn't the GPL? The GPL is not a social contract. The GPL is not the embodiment of the ethics of a community. It is not the constitution upon which the free software world is built. It is used as such by people at certain times. But in actuality, the GPL really is just a set of permissions to use software. And we build a lot of things on top of the GPL. We talk about the GPL as the protection of a certain type of freedom, as the arbiter of right or wrong. If you violate the GPL, you've done something bad. If you complied with the GPL, you've done something good. And we ask the GPL to play a role that goes beyond what we really ask copyright licenses to do in other realms. And that's sort of one of the main limitations of the GPL is that because it is just a copyright license, it's not a statement of ideals. It's a pragmatic legislation, if you will, of what you can do with a certain piece of software. And we're asking the GPL to play a role that it's not actually meant to play. And it's actually incapable of playing. And that leads to problems because people, when they think of the GPL as the embodiment of the right and the good, they don't all agree on what the right and the good is. So there's an awful lot of people who don't actually agree on what the GPL is. They believe that the GPL is a document that lets them do good things with free software. And to the extent they all agree on what is good, that's OK. But to the extent that they disagree, there are a lot of people running around with images in their mind of what the GPL says and what it lets them do that don't actually match the language in the license. And we've seen that many times with the old version of the license, there are lots of communities where people believe the GPL forbids commercial use of free software, for example. And that's not something the GPL has ever done. It's not something the GPL would ever do. But we've had clients with huge development communities, huge user communities who, on their web page, say, take, use this software. It's free software. Just don't make any money off of it. Don't use it for anything commercial. And they're shocked to find out that this isn't what the GPL says at all, but rather what they imagined would be a great software community, the values that they thought a software community should be built on. They assumed the GPL embodied these values. And this led to a lot of discord within the community between people who had read the GPL and people who hadn't. Because even the people who hadn't read the GPL were just surely convinced that the people who had read it were reading it wrong. And there was nothing you could do to change their minds, except take them to the FSF web page or take them to our offices and sit them down and try to explain it to them. And even then, people wanted to know, well, how could this be? How could the GPL, which I've put all my faith in, not satisfy that faith by being exactly what I wanted it to be? And that's a major problem and a limitation in what the GPL can do, could ever do. The GPL will never, ever be all things to all people. And it's a major limitation in having one license. It's one of the major costs of having one dominant license the way the GPL is. Now, I, of course, believe that having one dominant license is really good for inter-operation. But there are a lot of people who don't want to actually use the license as it's written. And they're going to get surprised when they find out that the license isn't just a embodiment of all their hopes and dreams. So that's the first major problem with the GPL is that it's not perfect. It can't actually play the role that a lot of people expect it to play. And that leads to a lot of problems within free software projects and communities. But even though the license isn't perfect, it's really, really good. It's so good that we've won. I mean, is there anyone left who doubts how good free software is, who thinks that the free software wave is going to end? Everyone is placing their bets in this area. People are investing in free software. Businesses are relying increasingly on free software. More and more projects spring up every day. And free software just gets better and better. And that happened partly because the license is really good, but mainly because we built an ecosystem. The FSF strategy of building basic tools that people could build on to create more and more free software worked. And we now have an economic rationale for using free software. We have software that is very sophisticated and very complex and very expensive to develop outside of free software environments. So you have a reason for businesses to use free software. And businesses use free software not because they love the ideals embodied in the GPL, not because they believe in the right and the good and the beautiful dream of everyone having free access to software. They use GPL software simply because it's efficient. Simply because there is no good business case for not using GPL software. And that's great. That incentivizes a whole lot of free software activity. It makes a lot of people happy. It produces a lot of new projects and a lot of support for developers. We're always in favor of software developers getting paid. We're always in favor of software developers being able to feed their families. The downside to all this is we've invited a whole bunch of people, a whole bunch of companies, into our community who don't share our values. We who work for organizations like the Software Freedom Law Center, we are starry-eyed idealists to a very large degree. We really believe in the value of spreading free software simply because we think free software is good for people. But that's not the value. That's not the ethic that drives for-profit individuals, small businesses, and larger corporations to get involved in what we're doing. So we have a whole bunch of stakeholders in our community. We have a whole bunch of people who have invested years of development, resources, business plans, going forward that don't share our values. And so when we drew a line and said we're going to start working on GPL v3, somebody had to say, well, what do we do about all these people who have jumped into our community but don't share our values? What do we do with the document going forward? To what degree do all of these new stakeholders with new values play a role in the formation of the GPL going forward? Do we owe them an obligation? What obligation do we owe them? I mean, could we, if we wanted to, create a GPL that drove the entire bus off a cliff and made it impossible for these businesses to operate? Yeah, we probably could have created a GPL that did that. And it would have been horrible for everybody. It might have been a very idealistic document that nobody used. But it wouldn't have done anybody any good. So for very pragmatic reasons, we didn't decide to drive the bus off the cliff. But we never really answered the question from an ethical standpoint. Do we owe any of these stakeholders anything in our community? What is the role in our community of people who are here not because they actually care about our values, but really because they care about spreading free software for their own profit-motivated purposes? So that's a limitation that we've never really even faced, I think, as a community. I don't really see a lot of people talking about it. I don't see a lot of people writing about it. And certainly no one's doing anything about it to try to start any sort of conversation. And I think at some point a conversation might be needed. We might need to actually figure out some way to talk about what is it that we owe each other? What do they owe us? And what do we owe them? And what does the partnership actually mean on a day-to-day basis and on a long-term basis? So another thing that the GPL is not is a trademark license. And this surprises a lot of people as well. It is a copyright license. It is a patent license. But the fact that it's not a trademark license means that it doesn't deal that well with attribution problems. It doesn't deal that well with one of the major pieces of producing a software project. You produce a software project. One of the first things you do is give it a name. If you've ever seen a project fork and then spend two months squabbling about what they're going to call the new fork, two months while no code is written and they just sort of throw names at each other, you'll understand how big a problem the brand of a project is. If you look at some of the larger free software projects that are out there right now, they have dedicated people within their projects who are brand managers. They're not coders. Large projects, some of them have people who have never written any code, never written any documentation. Their sole function within the project is to manage the brand, is to get people to know about their project, to get people to use their project, to market them. And because this is an increasingly large part of free software projects, it's interesting that the license just doesn't address it at all. And because people, as I said earlier, look to the license for what ethics we should have in our community, look to the license to embody the right path forward, no one really knows what to do about trademarks. The questions of who owns trademarks, how trademarks get managed during forks, what to do when people want to, the GNOME project, for example, has a foot. It's a little naked footprint that they would like people to be able to use, just so that they can say, we're a GNOME application or we're a GNOME compatible application. And they are struggling right now with crafting a trademark policy and a trademark license that will allow people to do this in some sort of rational way, without giving up their rights and interests in the trademark. But the biggest problem for them is not the mechanics of writing that trademark license. The biggest problem for them is the internal discussion of, well, what is it OK for people to do with our trademark? What do we want to allow people to do? What don't we want to allow our people to do? And they've sort of said, some of them have said that they wish that there was a standard free software trademark license, or at least a standard that people in the community sort of agreed on, that people in the community thought was at least a minimum baseline for responsible use of a trademark, for ethical use of a trademark. And I can see Wendy wants to jump in. Feel free to interrupt me. Back and forth is always good. OK, so I'm wondering whether you think that specifically delineated trademark licenses are necessary, or whether the defaults of trademark law, non-commercial law doesn't require, use doesn't require license, that referential use doesn't require a license, just need to be better specified so people understand that they don't need complex law. That's exactly right. And most of the time when I talk to projects, I tell them that they don't need a trademark license so much as they need a trademark policy. They need a clearly elucidated list of things that they think are OK. And that even if they had a trademark license, they were probably never going to go around beating anyone over the head with it. But because people look to the GPL as a license to tell them what is good and what is bad, we've habituated everyone to think that licenses are really good ways of telling people what is good and what is bad. And this is another major problem with the GPL, right? We took a document that was meant for doing specific things within a legal system. And we think that this is somehow a good way to communicate with a bunch of developers about what you should and should not be able to do with copyrights or trademarks. And I would argue that that's probably not true. You should not use legal language to communicate with people who are not lawyers. You shouldn't use legal documents to communicate with the general public. And so people always come to me asking me to write them a trademark license. I have not written any trademark licenses under slightly more careful analysis. It's always turned out that they don't actually need a trademark license. What they need is some pretty clear understanding within their community an agreement of what it is that people can do with the trademark and what they can't. And that, to me, says that because we've habituated people to look to licenses with the GPL, they're going to continue to do it. And that needs to be addressed. People need to understand that that's not actually the best way to communicate with people what you want them to do with your trademark. The license, in this case, served in some sense the same function as a Wikipedia page, namely a place at which the conversation had to happen. And the details had to be hammered out, including the legal details. And so people turn to it because that's where the work was done, not because people aren't reading legal language. Yeah, no. Is that about what happened? That's definitely very true. It's not that mistakes were made. And so we pushed the discussion into a legal document, and that was a bad thing to do. That's just an accident of history. That's where the discussion happened, so that's where the discussion is located. What I'm arguing now is that perhaps we should move the discussion elsewhere, that we should move off of the GPL as the place to put all our hopes and dreams, and that we shouldn't decide that GPL version four is the place to add a trademark provision, that we don't actually need to have all of our community discussions about what is the right way forward in a license, that we could actually have it in more guidelines. Because what actually happened, if you look at how software communities actually developed, because people had so little idea of what was really in the license, because so many people were operating with a sort of social view of what the GPL means, a common understanding of what the GPL means, the actual nitty gritty of the legal language was not the thing that nurtured all of free software. It was all the hard work that the FSF did talking about the license and producing really good software, and people deciding that they liked these values. It was actually a community coalescing around some ethics and ideals that produced free software, that produced the free software movement. The GPL was a huge piece of it, but the specific words in the GPL were not the thing that drove free software. It was the advocacy and the people coming together to do that advocacy and to produce really good software. Sure. I'm not a lawyer, so it just may not be the appropriate point to ask this question. But I want to ask about standards relative to your point about copyright versus trademark, one being in, one being out. I find that open source projects are trying to establish themselves as an alternative to the sort of formal standards process, which is too slow and dominated by certain class of vendors. Is this something that has come up as something that the GPL would or would not do? Is it on the copyright side or the trademark side that this would fit, in other words? Standards are not a piece of the GPL discussion so much. They're a big piece of the free software discussion, and the usual way they enter the discussion is people saying that the standards bodies don't serve their needs and they need to form new standards bodies or have informal standards or things like that. But it's not actually a GPL problem, and it's definitely not a trademark problem at this point. So I don't think that standards are going to become something that the GPL goes near anytime soon, but obviously standards are hugely important to free software. If you can't interoperate, you can't operate. And so we're probably going to see more and more standards activity that is not located in the GPL. And standards being what they are, you really do need very specific technical language to describe them. So that will not get located in licenses but in RFCs, I'm sure. So the GPL, it's a copyright license, it's a patent license, it's not a trademark license. And the one other thing it's not is it's not a document that reads itself. How many people here have read GPL version two? How many people here have read GPL version three? See, more than most places, but certainly fewer than have read version two. And so of the people who read version three, did you guys think it was easier to understand on one read than version two? Just casual, quick page through it? Anybody? I'm getting one squirrely hand in the back there. The individual pieces of the third were more clear, but thinking about the consequences of them was far more difficult. It's more clearly written, but it's far more complex. At least it was my impression. Version three is a longer license. It's got more pieces, and they interact in more complex ways than in version two. And that's just a function of living in a more complex world and asking the license to do a bigger job than the last license and to do it in more nuanced ways. So we've got a license that a lot of people are looking towards for their ideals and are looking towards for their sort of ground rules for their community. But it's a license that a lot of people have told me they can't actually understand when they read. And that might be a problem. That might be a problem when people who are trying to actually read the document and figure out what it means for their community as a baseline for good practices and ethical behavior when they probably don't understand it and might have violent disagreements over what the actual words mean when taken together. And I've heard that complaint from a number of not just individuals and free software development groups, but also small businesses. I've heard that complaint from lawyers for small businesses who have said, what does it all mean, and how can we find out exactly what is in the document? I've read the license, but I don't understand it, that kind of thing. I've also had a lot of people come up to me and ask me, hey, so, GPL version three, tell me about it. And they want the one paragraph encapsulation. And there is no one paragraph encapsulation. You can't really give it to them. And you couldn't really do that with version two either. All you could really do was point to the four freedoms. And that's sort of what I do today. I sort of give them the outline of what the four freedoms are and then tell them they should probably read the license. But that's going to be a problem going forward when people need to know what's in the license, but have a lack of ability to understand it. We're really lucky that we have organizations like the Software Freedom Law Center to help people understand the license, but frankly, we're a pretty small operation. And there's an awful lot of developers out there. We can't talk to all of them. The FSF is bigger than we are, but they can't talk to all of them either. So we're not really sure how the information is going to get to the people who need it. And I think that's going to be one of the biggest obstacles just going forward. Now the reasons why the document is longer, I think, are also really interesting. One of the primary reasons is that it's written at a higher level of abstraction than the last document. Rather than simply talk about what you can and cannot do with reference to American copyright law, in the effort to internationalize the document, we sort of described behavior that would give rise to legal implications under the copyright law of your country. And that's a more complex way of describing what we're doing. It's a more general way of describing permissions. And it makes the document easier to enforce and interpret under other legal regimes. But it doesn't make it easier to understand. And so you can sort of see the reasons why the document is longer and more complex. But the downside is this added complexity. So the big thing that the GPL did that I thought was really good was that it eased the path of compatibility. We have compatibility with licenses that, for a long time, people have been wanting to mix within software projects. Primarily, we've got increased compatibility with the Apache license. And Matt is our local GPL v3 license compatibility expert. So you can ask him all your details about it. But the Apache license, there's an awful lot of code under the Apache license. And the ability to get that code in projects together with GPL code is, I think, a huge boon to the free software community. We also have increased compatibility with a Ferro license code. And given the increased role that the, does everyone here know what the Ferro license is? So the Ferro license was a version of GPL version 2 that had an added clause, which basically said that if you deliver services to people using the software over a network, you had to give them a copy of the code, of the source code. In the past, you only had to give people copies of source code if you actually give them code, if you actually give them software. When you use Google, Google doesn't give you software. Google gives you a service. But they're running software. So if all of Google's search engine software were a Ferro license, it would be basically GPL'd with the additional requirement that you, as a search engine user of Google services, could download a copy of their code. And in a world where software is delivered as a service rather than as an application packaged for download, that's going to become more and more important. And the, the, the Ferro license, A-F-F-E-R-O. It's probably better called the Ferro GPL license, because it is a version of the GPL. The, the FSF, the FSF likes the Ferro license. We like the Ferro license. And there's a lot of interoperability language in GPL version three to make it easier to get code under the Ferro license. OK, yeah. That, that, the word, the word of Ferro got, the word of Ferro got in there. Because it was originally produced by a small, I guess, company named a Ferro. They wanted to create this additional requirement, and they contacted the FSF, and the FSF said, go for it. And so that's why you have the Ferro GPL. That was version two. And so now we have version three. And it's an exciting license. And if you're delivering services over the web, I encourage you all to use it. One of the limitations of the GPL is that obviously the Ferro license is more protective of freedom than the plain vanilla GPL version three. And yet it's not a default part of the license. Even though a lot of people within the free software community would like to get code from websites that they visit, would like to get at the PHP behind the pages that they see, in the same way that probably everyone here who can write any HTML, learned HTML by using View Source, we would like to be able to View Source for the back end of websites that we see as well. And we can't do that right now. But the Ferro, if it were widely adopted, would allow us to do that. And there are a lot of people who thought that should have gone into the GPL. Instead of being a separate license, it should have just been a requirement so that everyone could have that protection going forward. And it didn't make it in there. And this goes back to what I was saying earlier, about what do we owe the people who have built websites, who have built businesses around GPL software, who have relied on GPL software as users and as developers. It was ultimately decided that changing the rules at this stage of the game was probably not going to fly, that people would not appreciate it. And it was the equivalent of driving the bus off the cliff. There were a lot of users. I mean, when we got feedback from people who saw the first draft of the GPL, this was one of the things that caused the most consternation among people. So it didn't make it. But that's, again, that comes back to the complexity of having people in our community who don't necessarily share our values. So the last thing that I want to talk about in terms of the GPL and its limitations are the hard cases. There are pieces of the GPL that we didn't actually get to solve, because they were very difficult. And one of them is DRM. DRM is complicated. The access controls that authenticate users and authenticate applications to systems look a lot like security measures, in addition to protecting people from using content in ways that publishers don't want. These are the same sorts of systems that allow systems administrators to prevent people from downloading or using applications that they don't want, or require applications to authenticate to operating systems, for example, to make sure that you're getting official binaries from your distribution or whenever. And it's also, for example, the system that allows, say, networked games to ensure that everyone's client is a plain version of the client and not a hacked version of the client that allows you to always roll 20s or whatever it is. I mean, there's never going to be a system that allows you to authenticate a copy of a Britney Spears song that doesn't also allow you to authenticate a client. So the DRM discussion gets complicated in that it's really hard to carve out one without carving out the other. And I think we made a good case for it in that we talked about DRM in terms of copyright control mechanisms specifically under the DMCA. And I think there were people who wanted us to go further than that. I know there were people who wanted us to go further than that and prevent any use of GPL software for any sort of DRM application at all. And the limitations of reality were that we couldn't actually go as far as all that just because there were too many unintended consequences of capturing all of that activity. So we'll see how it all plays out. And I don't know whether it's going to be good enough, not good enough, or perhaps too good. But I think it's a really interesting area that we should probably watch, like whether or not that clause of the GPL gets any use will be a very interesting part of its enforcement going forward. So that's sort of my high level view of where I think limitations in the GPL are. And I think it's important for people to realize that a lot of these limitations are not drafting failures in the document. They're limitations inherent in using a document like the GPL to do way more than that document could ever do. Using legal documents to create communities is a very difficult thing to do. And I personally think that we should probably try to move beyond it at least a little bit. So that's my spiel on the GPL. And I will take questions from Wendy. Well, I'm just going to push you a little bit on the DRM discussion because, well, I agree that the technology is the same to authenticate a game client and to authenticate a Britney Spears purchase. I think we can make distinctions based on who's authorizing the unlocking of the technology and whether the user has knowingly bound him or herself into the system of control. So the game player might make the choice. I'd like to be part of a game network where no one can cheat. I'm going to agree to use this binary only client. The music purchaser hasn't generally made that choice or been given that choice. You've been given that choice, right? You could require that as a condition of purchase, right? Yeah, and I would love to see it as an explicit condition of purchase so that more people realize. You obviously haven't read the terms of service. That was what they were being forced into and could choose otherwise. Yeah, I mean, I think that you're right that there are differences between them and you could ask people to explicitly opt in. I think anyone trying to stop you from making off with a copy of Britney Spears would say that you have explicitly consented when you agreed to use the website. And it's in the terms of service and blah, blah, blah. But more importantly, writing that into the GPL, getting down to such a specific use case for a license that's used across so many different technologies and across so many different environments would be really difficult. I mean, we thought about it. I don't think that that's actually possible. I think that if you want a license that's going to be flexible and not tied to current technology too much, you're probably looking at more general language than that. Yeah, I'm not necessarily sure that the GPL would have been the right place to be more specific. I just really don't like the argument that bank security systems and DRM are the same thing, which is an argument that we hear a lot by people who really want to avoid restricting DRM, but say they're protecting all sorts of other assets. Don't get me wrong. I have as much hatred for DRM as Richard Stallman. I do think that there are a lot of similarities in technology and in crafting language where you're aiming at those technologies, we initially, I think in the very early drafts, wasn't there language about user access control? So when we started there, I think that's where the difficulties cropped up in terms of those sorts of use cases. And that's why we went around it by talking about specifically the DMCA of copyright access mechanisms. We've already actually run into, I mean, there are already some collisions where, you know, that in general, of course, we look at tibialization. It's not a great word. It's the one that Stallman uses. Because in fact, I think under the current system, you can cut off service. I mean, this is one of the sometimes you can. Sometimes, yeah. But these general cases of hardware lockdown, which are, of course, often used for strict people's access to the copyrighted works they've purchased and all kinds of other nefarious stuff. But we've run into, I've done some work with FCC regulations on software defined radio where you have these this. Which they say don't exist. Well, that's another story about what's actually being sold on the market right now. But you know, there's this federal mandate to sort of, at least under some readings, to lock down hardware. And it's made vendors very skittish about this kind of stuff. And so there are these cases where people are perhaps not entirely irrationally willing to limit some other freedoms for access to certain technologies. Well, I mean, I think the best example is probably the medical devices stuff that we talked about. There are, one of my co-workers, very sadly, is facing the possibility of having a pacemaker installed. And pacemakers are little computers. They run software. They don't actually have an operating system as such. But there is a chip, and it does have code, and it does have data, and it runs that code on that data. It's a computer. And the question of what kind of code is on that computer has come up. She's called a whole bunch of companies trying to find out, well, can I get a copy of that code? You know, you're going to put a computer in my heart. I'd like to see the code, just for my own peace of mind. And customer service is always very helpful. They always promise her, of course you can see that code. They don't know what code is. And so she calls around, they kick her upstairs, and finally someone says, no, we can't do that. And the scary thing is that the code that is on the device is changeable. Doctors can reflash your pacemaker. If they want to upgrade your pacemaker, they can change the code that's running on your heart. And you as a user, I mean, if somebody has the ability to change the code on a device that's on your heart, maybe you as a user want that ability. But doctors really don't want to give you that ability. And I know that you're all thinking, well, wow, I could overclock my heart. This would be really cool. But doctors don't want you to overclock your heart because they don't want you to cook your heart. And the access control mechanism that would allow a doctor to upgrade your pacemaker software, but not you as an end user, as a patient, is a lot like DRM in some respects. As it turns out, we suspect that there really isn't any access control mechanism. That anyone with the appropriate hardware could just come up to you, and if you would sit still long enough, communicate with your pacemaker. Maybe it's Bluetooth. We don't know. Maybe anyone with a Bluetooth keyboard. All I know is that this is a serious concern for people who want to put free software in medical devices is that a lot of companies who make medical devices are really scared at the notion of end users hacking their implantables. We're going to see more and more use of small computers in medical devices that get put in the human body. I mean, you can sort of imagine it with artificial limbs, for example. And users at some point are going to want to hack those devices. That's just human nature. Somebody out there is going to really want to grab his JTAG cable and go to work. But there's no insurance company on the planet that thinks this is a good idea. And there's, oh, there is. Let me, I can answer this first hand. Sure. The director of the devices division of the FDA. Many years ago, five, six years ago, when he was going out, as Clinton at point, I don't know who's there now, in a meeting with us, basically agreed with one of his staffers who saw open source software as a magnet for excellence. It's a direct quote. So whatever the insurance companies might or might not think, and I don't think it's at all clear what the liability issues are. Sure. Certainly the FDA, who matters most, because we are basically talking about the regulation of the object code, or the T-vote, that results from that open source software. Sure. They see this as a magnet for excellence. All the same reasons. And do they differentiate between licenses like the GPL, which have specific things to say about end user ability to change the code? Not five years ago. Sure, not five years ago. The problem that brought us there, and a related issue, has to do with the plugability problem, which is you don't really want to reflash your pacemaker. But you might want to have that particular, I like to swim two miles a day option, install from some other doctor. Right. And the ability of the FDA to regulate things that, whether they're open source or not, that then interact with each other is, again, a total mystery, but one that's getting to be increasingly relevant. Yeah, I think that's an open question. And it's only going to get more and more sticky as technology moves ahead much faster than we can write new licenses or create community ethics. I mean, there's going to be a lot of interesting edge cases in that field specifically. But all I want to say is that the open source software is a way of plugging components together. It's an alternative to having standards or buses by which these components would be joined. Traditionally, regulatory practice has depended on you have the isolation of a particular standard, you register, regulate your piece, and I register mine, and we're isolated by the standard. Open source software basically says, you don't have to do that. Just use the GPL. Yeah, I think it's an interesting area. I'm sorry, go ahead. The DRM stuff comes up in other instances as well. If you have a cell phone, there's E911 legislation that requires cell phones to be able to dial 911 and have the cell phone company be able to locate you so they can tell an ambulance or the police or the fire department, where did that call come from? Someone calls 911. They're complaining about a problem. They get cut off for some reason where they don't know where they are. They want to be able to send help. And I'm told, I haven't actually looked at the legislation specifically, that E911 requires not only that cell phone carriers and cell phone manufacturers make this functionality available, they also require that the user be unable to disable it. Now, this is mainly so that they can track you if they need to, whether you want them to or not, which sounds really scary. But presumably, it's for our good health. And that's why we need software to find radios. Yes, which is why we need software to find radios. Great disabled button on that. It's a powerful power button. Right, but a lot of people want to be able to use their cell phone as well. And the question is, can you deliver a GPL phone, a top to bottom GPL v3 phone in the United States without running a foul of E911 legislation? And if you can't, then that's a barrier to use of the GPL, which is actually seeing more and more uptake in the cell phone market provisionally, hopefully, sometime in the future. But more and more, I mean, I have this device. It's a N800. It's running GPL software, not top to bottom, but an awful lot of it is. And if they stuck a cell phone in it, that cell phone radio would not be running GPL software, partly because of the E911 legislation. And that has everything to do with the question of what happens when individual users under a copyright license run up against legislation, well, the legislation is always going to win that battle. But yeah, that's right. Just this example, is there a law against an individual user replacing the software in their phone? Or is it just that the, in other words, you're not drawing a clear distinction between the vendor selling unlocked phones in the sense of the word, and an individual user, or in our case, an individual doctor, reflashing the phone? If the individual user has the ability to replace the software to turn off the tracking, that would be a problem under the legislation. As it has been described to me again, I haven't actually read the legislation. The problem that I've seen, and I've looked at this one specifically, is when the legislation requires that the product be robust against user modification. So these robustness rules basically the ordinary user shouldn't be able to change the device to make its radio go outside the OCCI licensed band. For now. And so yeah, without specifically naming free software is forbidden, it all but says that. Yeah, I mean, that's one of the problems of free software, essentially not just being about freedom of software, but freedom for end users, at least as it's envisioned in the GPL, is going to run up against legislative restrictions that are, whether they're diabolical or not, are there in place for some sort of reason. And if we're going to insist as a community on freedom for everybody, there are probably going to be places where free software can't go, at least free software as envisioned by the GPL. And we're making a trade-off, right? We're making a trade-off between adoption of the GPL widely and more freedom for the people who can use the GPL in a narrower, slightly narrower sphere. Now, I think we're making the right trade-off by insisting on freedom at every turn. It's what the GPL has done all along, and I don't think that's the bad strategy. But the reality of it on the ground is that I've heard people within the cell phone industry expressing concern about whether or not they can use the GPL in cell phones for certain purposes. So sure. So if you want to have this, want to move the discussion away from the GPL itself, what are your thoughts about where it moves? How do we have that conversation? Yeah, and how do we get a version of the GPL that is both accurate and readable? OK, so the FSF has, on its website, I don't recall exactly where, a copy of the GPL with some parts italicized and some parts bolded and some explanatory material. And it tells you how GPL version 3 differs from GPL version 2. So if you have a good grounding in version 2, this is a really good document to read. It's sort of a shortcut to understanding where version 3 is. It's available on their website. I don't actually know where on their website. But gplv3.fsf.org is probably a good place to start. And in terms of the conversation that we have? In terms of where to have that conversation, that's a really good question. And it's sort of unfortunate that we don't actually have a place to have that conversation. I mean, if you think about all the various structures that exist within the free software environment, there is no central place where people go to have discussions like this. I mean, you could sort of talk about slash dot. But nobody really wants to have a slash dot level conversation about this. And they're the only ones who have all the different communities focused in one area in the big audience. So it's an open question. And if somebody were to start a project to sort of bring all these voices to the table, getting enough people to listen to them and participate in the discussion would be a major obstacle. So I'm not sure what the solution is there, but it would probably be a long-term project. It's not something that you could just announce and do overnight. Anyway, that's what I have to say about the GPL. And I came here today, I was talking to Amar on the phone last night, yesterday afternoon, and I didn't know what I was gonna talk about today. I thought I might talk about the GPL and I thought I might also talk about technology in internet radio. And I decided not to talk about internet radio, even though it seems like, from talking to some people before the discussion, that there's a lot of interest in it here. So I brought some copies of a small piece I wrote on the subject, which you guys can read, and you're welcome to email me and tell me you hate it, tell me you love it. So thank you, everyone, for coming and hearing what I have to say. Thank you. Mr. Clark. Thank you. Thank you.