Contribute and be proud of defending Software Freedom from monopolist control! We are intervening in the Apple vs. EC case: Find out more.

GPLv3 logo

Transcript of Eben Moglen at the 3nd international GPLv3 conference; 22nd June 2006

See our GPLv3 project page for information on how to participate. And you may be interested in our list of transcripts on GPLv3 and free software licences.

The following is a transcript of Eben Moglen's presentation made at the third international GPLv3 conference, organised by FSFE in Barcelona. The conference page has materials from the other presentations.

Please support work such as this by joining the Fellowship of FSFE, and by encouraging others to do so. Transcription of this presentation was undertaken by Giacomo Poderi and Cristian Rigamonti with checking by Ciaran O'Riordan. The video was made and processed by Sean Daly.

Eben Moglen is General Counsel of the Free Software Foundation (a sister organisation of FSFE), and is the Chairman of Software Freedom Law Centre.

Video and audio recordings are available:

The speech was made in English.

Presentation sections

  1. The presentation
  2. First: Bill Gates has surrendered
  3. Where are we?
  4. The public process
  5. How the committees discuss
  6. How the public comments
  7. An example high-value comment
  8. A lot has been rewritten
  9. Internationalisation
  10. Example of bridging legal system differences
  11. Propagating and conveying software
  12. No more reliance on "derivative works"
  13. Fixing licence proliferation
  14. LGPL, like merging electronic weak
  15. The patents issue
  16. The DRM issue - it's all about software
  17. What happens after the process

Questions and Answers session sections

The presentation

[0h01m08s]

Eben Moglen: So, I'm Eben Moglen, I'm a lawyer who works on the production of freedom, mostly with the makers and distributors of Free Software and mostly at the moment for one client, whom you know.

Because Richard was kind enough to go through the state of substantive affairs in the drafting of GPL3, I want to take time this afternoon, before answering your questions, to give some context for the effort that he described in detail.

[0h01m58s]

First I think I owe you as citizens of the Free Republic some more precise information about the process and its state at the present moment, and then I'd like to talk a little bit about the larger aspects of our life in the Free Republic into which GPL3 fits.

[Section: First: Bill Gates has surrendered]

But the first thing I think I ought to do, is to give you some news: Mr. Gates has surrendered.

[laughter]

And before we get all back to fighting again among ourselves, we ought at least to recognize the meaning of that step. In the United States when a major corporate executive resigns or retires there is a mandatory process of talking calm to the markets. You don't resign on less than two years notice, so as to give the feeling of orderly transition and complete control over circumstances. If, for example, the chief executive officer of the Enron corporation resigns suddenly on no notice, then that's a sign that immense trouble is immediately ahead as in that case bankruptcy was by six weeks.

[0h03m35s]

So for Mr. Gates to have announced that he is retiring from active service to Microsoft on two years notice, is the minimum length of time, that could have been chosen without sending the signal: Microsoft is in an acute short term trouble.

It isn't "My God! Everything is great, he's giving two years notice". It's "That's the least that could have been done".

However, this particular retirement was announced to occur approximately five months before the most crucial product release in ten years. So anything that leads you to believe that this is normal, ordinary, "Mr. Gates promise Mr. Allen and Mr. Ballmer that he would stay at Microsoft until at least fifty, which he now is" - all of that talk is irrelevant. Mr Gates last week announced his retirement on the shortest possible timeline that wouldn't spook the markets. A resignation announced six months before the shipment of the most important product in ten years. I am characterising that fairly When I characterise it as personal surrender.

Even better: Mr. Gates has surrendered and is leaving the company to Mr. Ballmer. Better news for us could not have been imagined. So, that's where they are.

[0h05m23s]

[Section: Where are we?]

Where are we?

We are on schedule with the most important work we have done, in fif***** years. Anybody who thinks that Mr. Stallman's retirement is about to be announced wasn't here this morning. He's fifty four, Mr. gates is fifty, Mr. Gates has quit and left the field to *****er men, like Mr. Stallman. So here we are, in the middle of the GPL3 process which, as to timing at any rate, seems to be going on at the right moment.

As you all know we released a first discussion draft in January at MIT. We said: "it's a discussion draft, we are listening carefully to what people may want to say about it" and we convened a large and, admittedly, somewhat complicated process for listening to what people had to say.

[0h06m26s]

The first thing that turned out to be on everybody's mind was that GPL version 2 was perfect. I know that it was perfect, because everything we wanted to change they told us we shouldn't change at all. And the implication therefore was: we have been living with an absolutely perfect licence, every single change in which was a bad idea. I wasn't absolutely surprised about this. Human beings have a certain resistance to change, and I wasn't surprised that everything we proposed changing turned out they really loved in the old version.

But it is good to know, at any rate, that GPL2 is perfect. So if anything goes wrong, and we don't do GPL3, the World will still have a perfect licence with which to work.

[0h07m10s]

I actually think, if I might dare to say so, after fif***** years of working with it carefully that GPL2 can be improved, and we are improving it, but it's nice to know that the world at large have come to regard it with such fondness before we start to change it.

Within a couple of weeks, after our announcement of the draft and the process for discussing it, I began to hear rumors that we weren't really going to listen to anybody.

Apparently we had gone to enormous amounts of works to create a worldwide process for listening to everybody solely as a sham; actually I've heard: "Mr. Stallman is going to make the decisions all by himself", every once in a while I even heard people saying that I was going to make decision by myself, which if they knew Mr. Stallman they wouldn't say.

[laughter]

So in any event for a little while there, it looked, I suppose, to some people as though this whole grand effort to get everybody else's advice and wisdom involved, was somehow just a game we were playing. Which is, from my point of view, too bad that it's not true. If this were a game we were playing about taking everybody's opinion seriously I could get a lot more sleep, but unfortunately it's not that way.

We actually set out to listen to everybody as carefully as we could in a context in which each kind of commentator would feel comfortable making his and her kinds of comments and the commenting on behalf of their organisations. And we've been listening extremely hard for four and half months and there's a great deal that we have learned in the course of that time.

[0h08m46s]

[Section: The public process]

I want to talk just a little bit then about that process which the world at large has seen, but only one side of.

We made discussion committees by asking those people who joined us in Cambridge for the launch to join into committees, organized around types of social role in the process:

  1. vendors of commercial versions of Free Software or hardware that works with Free Software
  2. leadership of Free Software projects that do use GPL, that don't use GPL
  3. lawyers on behalf of large users, corporate officials working on behalf of corporations that are large users, public servants who are interested in public sector adoption
  4. individual hackers with substantial street cred whose views might be influential in helping other people to consider the licence either as one to use or as one to approve, and so on.

[Section: How the committees discuss]

Those committees meet in different ways, one committee consisting entirely of hackers prefers to work almost exclusively by IRC, one group consisting very largely of persons employed by major vendors tends to use large conference calls, parties use the forms of interaction which are comfortable and convenient for them, and we attempt to provide an environment in which everybody's comments can be heard and thought about with a sense of equality among participants.

[0h10m23s]

[Section: How the public comments]

As Richard said, we created a little web services code that didn't exist before. We needed an application for allowing a very large number of people to attach comments to a single document and for the world at large to be able to see in an intuitive way, reading the text, where the comments concentrated in what level of intensity and how to click through there to get to other people's comments quickly on particular areas of interest or concern.

I actually think that software did a pretty good job. It reduced the volume of comments, somewhat, at the cost of, or at the benefit of, greatly increased specificity in comments, because the Stet system made people go and highlight an area of text and say: "this is what I'm talking about and this is what I have to say" the quality of comments was very high. And the number was rendered more manageable because people didn't file, essentially, comment spam saying, I hate the whole thing or I love the whole thing, they actually went and found a place where they had something to say and said it. We had about a thousand non trivial comments from third parties not involved in the discussion process in any way, just people who used the system to register commentary of importance to them.

In that thousand, every one of which was reviewed by somebody working on the process with me, with Richard, with David Turner, with Richard Fontana. In that process of looking at every single one of those comments I will say that at least one, maybe two or three, really important issues were seen for the first time, by somebody unaffiliated with our process in formal terms who simply had a damn good point to make which needed to be taken into account.

[0h12m21s]

[Section: An example high-value comment]

Let me give you one example of an important issue unearthed in the public comment process which will have an effect on the second discussion draft of the licence.

As you all know, since the beginning of the licence it has concentrated a great deal of attention on distribution as the moment at which obligations attached to your role in the free community. The assumption has always been that distribution is a one way activity. Somebody distributes, and a user receives. If binaries are being distributed, in particular, the binary distributor has an obligation to make source code available to a receiving user.

All of this made a great deal of sense in 1991 and it probably still makes great deal of sense in 2006, but. Consider the situation of an iso image of a CD containing binaries of Free Software material under GPL being distributed among persons by BitTorrent. If you think about that scenario for just a moment you will see that in the torrent are a bunch of people "distributing binaries" who don't think of themselves as distributors, who think of themselves as users receiving the binary who have no source code and who haven't even got all the binary, yet. So that even if source code will later be part of the same torrent they're not guaranteed to have it and they can't pass it along.

Under GPL2, an unaffiliated commentator said to us, this would be infringing activity, why don't you do something about that in GPL3 ?

We looked at it and said, "you know, you're right". That's a major problem. That is to say, it's not all that hard to fix, but it's a major problem not to have seen it, "we are glad you brought it up".

And we will make a few very narrow changes in the second discussion draft of GPL3 to allow for the fact that peer-to-peer models of distribution may involve ancillary transmission by people who would traditionally have been thought as receivers and they shouldn't have any obligations under the licence, because they are not actually engaged, in the sense we used to mean, in 'distribution'.

[0h14m55s]

That's an example an important technical issue in the changing, use and context of the licence, raised by somebody from outside the process and, as I said, there are probably two or three more of equivalently important scope that I could point to. That kind of commentary is worth its weight in gold. It is because of that kind of commentary from unaffiliated individuals that the entire process is worth maintaining, cost-what-it-will, to maintain it, to read its output, to think about every word it says, that's exactly what we were hoping for, that's exactly what we are getting.

I have enormous respect for those people around the world, who've decided to take the time and trouble to make comments on this licence. They are doing a bang up job. As somebody who has to think about this all day every day, right now, I am immensely pleased that smart people would be willing voluntarily to take some time on it, and the quality of what we have gotten in every way demonstrates the power of the multiplied hive mind. We are getting great work out of people.

[0h16m02s]

This is true of the committees. [inaudible short sentence] Committee members have spent dozens of hours, in some cases hundreds of hours, working on this licence. They have spent a lot of time communicating, they have spent a lot of time refining ideas, they have spent lot of time coming up with improvements or suggestions or possible changes in language. They have been willing to do what in inter-corporate negotiations is often hard to do, they have come out and said what they wanted and said what they needed clearly and candidly, and I'm very grateful to them for doing that.

It has made it possible for us to understand what it would take to provide a licence whose users - commercial as well as non commercial - would feel themselves protected in their rights. And that's a powerful and important task that we set to perform and that we could not perform without all the effort and time and thinking that has gone into it on other people's parts.

[0h17m01s]

I believe that the first several months of discussion of the first discussion draft have been beneficial to the quality of the draft in every way.

[Section: A lot has been rewritten]

As you will see when it is released there are few articles that have not been substantially rewritten. I am still puzzled by people's belief that we weren't going to listen. I know that if I could release right this minute by throwing up on the screen a strike out and bold copy of the new draft, showing everything removed and everything inserted, that it would be harder for people to take seriously this claim that we weren't listening. They would see their language coming out and language going in that they would recognizes as directly responsive to public comment and to suggestion. In a few weeks that's what people will see, and I believe that that will allow us to enter the second discussion period with much more confidence on the world's part that the work that they are doing to help us improve the licence is being gratefully and thoughtfully received.

[0h18m15s]

[Section: Internationalisation]

In the process of revising the licence's second discussion draft, we have done more work on what I think of as the fundamental challenge in internationalisation. A question was asked this morning about translation, and translation is unquestionably one possible way of internationalising a licence. But translation is, in my judgement, at best, at best, a second best way of internationalising the licence, and, for the reasons that Richard offered this morning, a somewhat dangerous way.

The best way of internationalising the licence, in my judgement, is to look as firmly as possible to the work done more than a hundred and ten years ago in the making of the Berne Convention. An attempt to harmonise global copyright law in the World is now more than a hundred and ten years old. The United States was late to come to Berne as a full participant in the harmonising of copyright law. That's the good part of the thing called The Digital Millenium Copyrights Act, there is a good part, trust me, not much but a little.

And given that we are all now existing in a Berne universe, the right way of internationalising is to come as close to Berne Convention mechanisms as possible. In order to do that the one step really needed is to get out of the licence dependence on any single legal system's vocabulary or unique conceptions. And so our effort has been to genericise the way the licence does its work.

[Section: Example of bridging legal system differences]

I'll give you an example which came up in a conversation I had with Carlo Piana over lunch. One defect in the licence as seen in some countries of the world under GPL2 was that GPL2 stated that it was a perpetual licence. German speaking lawyers, for example, working in Germany, and that's only one example, said "you know, our legal system really doesn't like that much in fact, in order to be valid licences need to state a term". "Well, alright", we said, we understand that there are some legal systems around the world in which stating a term may be necessary or desirable in order to make a good licence.

There are also legal systems around the world, we work in one in the United States, where making licences perpetual is the right idea, because having non exclusive licences good only for a term is an invitation to trouble.

One way you could go at this is to translate the licence so that it says one thing in German and another thing in English, but as Mr. Stallman pointed out this morning, that's exactly the danger.

[0h21m19s]

Otherwise you appear to be put to a bad choice, choose one or choose the other and somebody will be unhappy with you. So the goal in modifying the licence is to do neither and still achieve the result needed by both. As you noticed, the first discussion draft said: "GPL now is a licence for a term, and the term is the length of copyright on the underlying program". Perpetual so far as the copyright is, and no more perpetual than that, and can be stated as a fixed term reduceable to a number at any given moment.

Accordingly, I hope and believe, based on the legal advice we've taken from lawyers around the world, who are helping us with this matter, we have succeeded in finessing a problem where we might have been thought to face a choice, between something that would work in system one and something that would work in in system two, but not both.

[0h22m12s]

The second discussion draft of the licence goes still further in that path. As Richard mentioned this morning, we have once again attempted to take vocabulary and consolidate it, so that there is a set of terms unique to GPL, all of which can be defined by relation to underlying copyright law in each individual copyright system.

[Section: Propagating and conveying software]

So "propagation" is doing anything with the program that requires local permission on the copyright law, except running it or privately modifying it.

"Conveying" it, is propagating the program in a way that makes it possible for others to make or receive copies. By using language which is not US copyright language, or anybody else's copyright language, but by stating only the relationship between licence terms of art and states of fact in the world, we think we can get rid of a large number of problems arising out of national law.

[Section: No more reliance on "derivative works"]

I will say, therefore, the second discussion draft of GPL3 will not have any longer any dependency on the American copyright law idea of a "derivative work". Accordingly we will be able to avoid another fif***** years of complaining from lawyers who quite justifiably thought the reliance on a US centric view of the derivative work was a major source of uncertainty and uneasiness. They may think that what we have done instead is not a perfect improvement, but I feel certain that there will at least be the recognition that it's a useful development.

[0h23m54s]

Internationalisation of the licence, as I said at the beginning, in Cambridge, goes along with the other major structural change that we wanted to make. Which Richard described this morning as we described it in January as enhanced compatibility.

[Section: Fixing licence proliferation]

I want to go back over that ground a little bit for you, in order to put the context a little more deeply around it. After Richard's discussion of enhanced compatibility this morning, a question was asked in the question period: "what about licence proliferation?"

Our answer to licence proliferation, I think, I want to expand a little bit on Richard's remarks, is very straight forward now: use GPL, add the permissions that you need, if the restrictions you would have put in a different licence are restrictions compatible with GPL under the section 7, you need not make any new licence. If you don't like copyleft we can do nothing for you, but if you don't like copyleft then you should recognise licence proliferation as the necessary consequence of the non-copyleft licensing you prefer to do. In other words, ask yourselves: "why has there been licence proliferation?". There has not been proliferation of copyleft licences, or rather there has been very limited proliferation of copyleft licences. There are two or three copyleft licences in the world with any significant mindshare.

Permissive licences have multiplied all out of keeping. But it is precisely because permissive licensing has as a necessary consequence the proliferation of licences, that we are where we are.

Parties who wanted to licence in non proliferating ways had too few choices, when there were a small number of copyleft licences and they all worked in pretty much one way only.

[0h26m05s]

GPL's new section 7, the enhanced compatibility provision, will allow copyleft licensing to be used more flexibly while still striking the balance in the direction of commercial usability for businesses and clarity for individual developers who want to know what licence to apply to their code. I think it makes an enormous contribution if people care to use it in that way, that's up to everybody else.

We will have laid down an infrastructure for licence interoperation which we believe, if used fully, will have a very substantial effect in reducing proliferation over the next few years.

[Section: LGPL, like merging electronic weak]

Some comment was made this morning about the LGPL as a section 7 additional permission on top of GPL. I would have left that subject to be disclosed when we show you the drafts in a couple of weeks, but as the subject has come up, let me just say one little thing about it. That move does not change the behaviour of LGPL at all, as Richard told you this morning. We adopted a "no functional change" constraint in the reshaping of the licence. What it does is to allow LGPL to serve as a good example of the additional permissions structure for GPL and how powerful it is. By expressing LGPL as just an additional permission on top of GPL we simplify our licensing landscape drastically.

It's like for physics getting rid of a force, right? We just unified electro-weak, ok? The grand unified field theory still escapes us until the document licences too are just additional permissions on top of GPL and I don't know how we ought to get there. That's gravity, it's really hard. But the physics has gotten simpler and, I think, that alone is valuable. The other thing is that the new LGPL expressed as permission will show people how to use permissions to make quite sophisticated licences very simply, and we think that also is all to the good because such licences do not proliferate, they reconvert to GPL at the instance of modifiers which is a valuable and useful property to have.

[0h28m43s]

That's all I want to say about the technical material in the licence, except to talk about "issues". As we said at the outset there are two or three dominant issues that remain to be talked out in order to get to a consensus about the licence.

One issue, or one group of issues has to do with the handling of patent rights. The other issue has to do with the control over software in the interest of entertainment security. I want to talk very briefly about both of those because Richard talked at length and I don't want to redo anything he said. So I'm gonna try to add only, and in questions you will take up what you think I have been too light on.

[Section: The patents issue]

With respect to patents, the issue has changed radically in the last couple of years. GPL version 2 is a licence that says: "There's a terrible patent problem, if you don't do something about it, all hell is going to break loose sooner or later". 'Sooner or later' has come and gone and everybody knows that all hell has broken loose.

The fundamental question now for everybody is how to invent around the patent system, through the licence, so as to achieve optimal goals for several different classes of parties. Parties who have large patents portfolios and want to contribute to the Free Software world and GPL in particular, parties who are afraid of other people's patents because they make revenues that could be the source of claims for royalties, and parties who want to develop software and are afraid of patents, not because they have commercially and significant revenue streams, but because they are afraid of the chilling effect of announcements of patents or patent licences or patent lawsuits as a way of politically inhibiting people from making desirable Free Software.

Those are quite distinct groups, some parties, that is parties with large patent portfolios also having significant revenue streams, appear more than once in the interest group analysis of the patent problems. Each of those major industrial parties with large patent portfolios has well developed patent policies of its own and well developed positions in negotiations with other parties in the industry concerning its patent portfolios. We, although we have a certain sophistication in those conversations, having sat in them for years, have no patents and no clout in that world, usually.

[0h31m39s]

Right now we have some clout, because we make a licence people need. But we cannot overuse that clout, we must make a licence that people need as much as they need the one we are changing.

The discussion about patents has been very complicated, very thick, sometimes rather heated, though those parties who have tried to apply fire to the soles of my feet have been kind enough to do it privately instead of publicly, still there has been a great deal of attempted torture. I have responded as gently as I possibly could but sometimes I have said harsh words back, too, for which I am now publicly stating that I don't regret it at all. Bargaining is bargaining; we will be done soon.

[0h32m29s]

I believe that great progress has been made among the patent holders in understanding what they need, what they need collectively, as well as individually and what represents good outcomes from the Free Software Foundation position in this matter. We have considered a number of changes to the statements contained in the first discussion draft about patents, we have listened exceedingly carefully to everybody who had anything to say on that subject, whether they were patent holders or not, and we think we have proposals to make, in the near future, that will be very helpful in resolving these issues. Once those issues are resolved we need movement on other questions too.

[0h33m13s]

[Section: The DRM issue - it's all about software]

Which brings us to the subject of Richard's major address this morning, the problem of Digital Restrictions over code.

People have tended to think of this as though Richard was trying to cram down their throat a replacement word for what they were doing. Companies that make entertainment content, think they have digital rights, and that they should be allowed to manage their digital rights.

I have no stake in that, in one way or another, as private citizen I have real strong feelings about that. But my role right now is to be a lawyer for the Free Software movement, and as a lawyer for the Free Software movement I have nothing to do with music and nothing to do with movies. I don't represent the Free Music Foundation, I don't represent the Free Movie Foundation. I represent people care about making Free Software, and for them Mr. Stallman is quite right. What are Digital Rights Management issues to the entertainment companies are simply Restrictions Management to us.

Because what is restricted is the right to make our software the way we make our software. That restriction is set to follow from somebody's use of his own rights over his own copyrighted works, music and movies, but that means nothing to us. Not because we think there shouldn't be copyright on music or movies or because we think sharing should be free, whatever it is. That's all irrelevant.

You will hear a lot of talk about DRM in which it is suggested that we are somehow trying to use a software licence to address non software issues. Nothing could be further from the truth: we are addressing a software issue.

Is GPL'd software being distributed in a way which would be illegal, if the licence were being evaded, and which is solely regarded as legal because the licence evasion is technically implemented.

[0h35m20s]

If I sell you a computer containing software under GPL, let's say I make a digital video recorder, let's say I call it TiVo. And I sell it to you and I say: "there's GPL'd software in here, here is the source code. The only thing is, if you modify this software inside my box I will cut your service off". That would be a straightforward violation of GPL. You would be adding a condition to the licence and you couldn't do it.

Now why it's such a big deal, that everybody should jump up and down and yell and scream and carry on if we say: "oh! and by the way, what you are not allowed to do by illegally modifying the licence, you are not allowed to do by modifying the hardware so that the licence can be evaded safely!". It's a straightforward proposition: you can't do it this way and you can't do it that way either, why are you trying to evade the licence?

"Well we are very concerned about copyright on music and movies" - Don't give me that. That's like any other reason for evading the licence. "Well, we're very interested in making money Mr. Stallman, and if we break your licence we can make more money!" - Sorry that's not a reason.

So what we are talking about is not something about why we are mad at Hollywood or why we are mad at the music factories, that's another discussion.

We are saying that the licence should prohibit technical means of evasion of its rules, with the same clarity that it prohibits legal evasion of its rules, that's all.

[0h36m56s]

On that subject we do not believe that there is a lot of room for principled argument going the other way, but such room as there is we are open to listening to.

Richard emphasised today that if this attempt to inhibit evasion of the licence came in the form of a refusal to permit use, it would be wrong to do. Some of the language we published in January lead people to conclude we were imposing use restrictions in order to deal with a mode of evasion of the licence by people seeking security for entertainment.

To the extent that that language caused to people to feel that that's what we were doing, it was plainly badly chosen language and it will be replaced in GPL3's second discussion draft.

[0h37m59s]

Richard also said this morning that we have been working against the background of a moving target in Europe. As European domestic implementations of EUCD threw up new legal restrictions that we might have to confront and deal with.

At the present moment we believe that the next draft will contain language successfully repelling the force of both of US DMCA and the EUCD, without employing use restrictions or other inappropriate technology within the context of the licence. We are continuing to work with lawyers around the world to see to it that the final draft will also act to get in the way of anti-circumvention law written elsewhere in the world using other legal concepts.

The goal is to make it possible for people who get GPL'd software to exercise the rights promised to them by the licence.

[0h39m04s]

Audience member: Excuse me Mr. Moglen would you like speaking a mite slower so we can...

Moglen: Sure, I apologize. I also apologize for using English. I appreciate that that may be difficult for some people.

To go back over that point, I wanted to say that we will use, as best we can, other legal advice we have taken to ensure that the final version of the licence, works against other anti-circumvention law around the world not merely EUCD and US DMCA statutes but other legal regulations elsewhere in the world as well.

[0h39m54s]

[Section: What happens after the process]

I want to wrap up and take questions for the rest of the time, I suspect there are a lot of questions. Let me just say something, because for me this is the half way point in a long process, let me say something about what will happen when it's over. You may be less concerned about that than I am, half way through this process I am very eager to think about the day when it will be over.

When it will be over, I believe, probably some time next January, we will have made, by international collaboration, the rules that govern a system of intellectual creativity, also the rules that govern a multi billion dollar industry, also the constitution of a social and political movement called the Free Software movement. We will have adjusted the language of all three documents inside one piece of paper. The rules that allow individuals and students and classes and schools around the world to experiment freely with all forms of computer knowledge. The rules which allow multi billion dollar businesses to operate in stable and unified ways. At a time when the leading alternate model of how to make and distribute software is visibly bankrupt and incapable of performing.

And we will have made clear what the constitution of the Free Software movement is, for the next little while, by stating what its goals are and how it proposes to achieve them.

[0h42m01s]

We will all have done all these things, tens of thousands of people around the world will have collaborated to do those three things, this year.

The largest, most deeply funded monopoly in the history of the world, will have lost its CEO, fired a bunch of its subordinate executives, and produced, instead of the best release in ten years, a total commercial and technical failure.

Proprietary software will have visibly collapsed, cut off at the knees by its own inefficiency.

We, you and I and everybody else doing this, will have succeeded in negotiating the rules of the industry for the next decade by open collaboration.

At the end of that process, politics and government and society and economics will notice that discrepancy. They will notice that discrepancy because the sound of the failure of Microsoft will be much quieter than you think, the sound of the failure of Microsoft will concentrate a lot of other people's attention because Microsoft won't be speaking so loudly anymore.

On the contrary, people will be saying to themselves: "They used to have a billion dollars a month in profit! And they used to spend that billion dollars a month making propaganda about how their way was the only right way. Funny how they don't have that profit anymore. Funny how they can't spend all that dough on propaganda anymore. Funny how much simpler it seems now, when they're not yammering so loudly. By God! Now we can hear ourselves think." And in the middle of this quiet we notice that theirs was a bad way of doing business.

That's where we are going to be when this is over. We are going to have renewed our fundamental infrastructure and made it much stronger and much more flexible. The world's leading alternative approach will have died on the field of battle chosen by itself, having failed to make its own products well enough to convince the users, it had already trapped, to use them.

That's going to be quite a shocking comparison to those who didn't know what we were up to as clearly as we did.

[0h44m58s]

That's all I have to say. Let me take your questions, thanks very much.

[applause]

[45m 00s]

Q1: I'd like to ask you to explain further about the difference between added restrictions and added permissions, which, at least to me, has been a bit unclear in the explanation of Mr. Stallman, and yours was a bit cursory, especially as it amounts to patents. Added permissions, as "you can use patents in our portfolio", and added restrictions, as "you can't use, or 'disseminate' is the word, our software if you initiate a lawsuit".

The second question is: if you've talked to people like the Debian project, who are seemingly even stricter than the Free Software Foundation in some of the requisites for what they call freedom in licences.

It's a bit of a long question, with three parts, I realise.

[45m 50s]

A: Yeah, they're actually quite different, let me try and take them as different questions, I'll speak to them differently as quickly as I can.

First, I thought that one of the disadvantages of the way section 7 was released in the first discussion draft, was that it attempted to treat permissions and requirements together, in one unified set of drafting. I had an alternate drafting proposal, which treated permissions as one sub-subject, and requirements as another; and in the second discussion draft we adopt that drafting instead. The consequence is to make the differences among permissions and requirements easier to understand and to follow.

[46:40]

Rather than talk my way through language that we'll have out on the street in a little bit, let me just put it in the narrowest way. Permissions are an essentially open set, you can permit pretty much anything you want in addition to the GPL, but permissions are removable. Additional requirements must be one of a specific set of possible additional requirements; the set is closed, not open. Those requirements are not removable from the code which bears those requirements, instead you have to point each recipient of the code clearly to which additional requirements apply to which portions of the code. So if you adopt something out an Apache licensed program, there is a patent retaliation clause in the Apache licence, therefore you have some code in your GPL'd program which has a patent retaliation provision applying to it, and you have to indicate clearly to the user what code that is.

That's the nature of the difference between permissions and requirements. You asked specifically about the use of the additional permissions section to deal with patents: we don't rely upon the additional permissions section to deal with patent rights. Whatever is going to be said about how patent rights are allocated between people making, people distributing, and people using the code, will be set in the base GPL, in section 11; and how it is set will be carefully determined after discussion with all those people in the world who care about that, but there we're looking for a unique consensus solution.

[48m 30s]

[interruption from the audience]

With respect to permissions, grants of patent allocations, non assertion of patents, whatever it is that is the technical material, the way you know that you get what you need to practice the freedoms in the licence, will not be as a result of additional terms. That will be in the base terms.

Patent retaliation clauses, however, go in that category of additional requirements sometimes permitted, and there the principle was stated in the first discussion draft of the licence: patent retaliation clauses are OK if they are fully defensive in character; if their purpose is solely to repel patent lawsuits, and not to take other kinds of advantages using patents. The GPL will not contain any such clause; it will merely say it's compatible with licences containing such clause.

So, to pick up your last question, there will be no issue with Debian: if a licence out there in the world is also a free licence under the Debian Free Software Guidelines, then that licence, to the extent GPL permits, may be compatible with GPL, but the freedom or unfreedom of that licence can be judged by looking at that licence standing alone under DFSG; no licence becomes GPL-compatible, which isn't also capable of being compatible with the Debian Free Software Guidelines, and we won't be saying "Licence X is compatible with GPL", we'll be saying "Here's a term; GPL can use that term; if GPL accepts code under that term, that term stays on that code, not under GPL, but under the other requirements set applicable to that code. Use all the code under the GPL, if you take that piece out, remember to pass it along with the patent retaliation clause on it". That's all.

[50m 35s]

Q2: Do you expect that GPLv3 with be compliant with the Debian Free Software Guidelines? And is it a goal for the next...

[50m 50s]

Well, I haven't any reason to suspect that there will be an incompatibility; I haven't any reason to believe that there is a chance of an incompatibility; it hasn't been my primary concern, because it hasn't seemed to me to be even the beginning of an issue; if it were suggested somewhere along the way by our own evaluation, or by somebody else's, that there was a real problem with the Debian Free Software Guidelines we would address it.

[51m 20s]

Q3: I have another question concerning the DRM clause. There is one wording that says that covered software cannot be an effective measure against... er in the sense of the DCMA. Do you expect this clause to be effective?

[51m 45s]

I expect that US courts will be instructed on the intention of the licensor to reject the features of DMCA as it applies to GPL software. I expect the United States courts to listen closely to statements of the licensor's intent, because under US copyright law it is the licensor's intent which normatively determines the content of licence.

That language was not meant to deal with EUCD arrangements, which are different; we will employ a different strategy for dealing with the EUCD, which is not present in discussion draft 1, and which will see its first public exposure in discussion draft 2, at which point we will then begin discussing with European lawyers, businesses and parties, what they think of that proposal.

[53m 00s]

Q4: [question asked without microphone]

[53m 05s]

Yeah, the question was asked offline whether I would now disclose what that's going to be; I have a 95% certainty that I know what the language is, but might I point out that there are two countries presently implementing EUCD right under our noses, and I reserve the right to make further changes in light of those legal changes if I need to in the next several weeks. The reasons for not pre-disclosing things is that I have an obligation of professional caution, and my client as you know has a very deep belief in the importance of avoiding mistakes that could permanently harmful to freedom. If this were the right date to disclose those terms, we'd disclose those terms today.

[53m 50s]

Q5: On the DRM front, is the invasion of privacy wording one of the one wordings that's going to be removed from this draft?

[53m 55s]

Yeah, I expect that language to be gone; I've said that in public on several locations. I thought myself as one lawyer that there was a useful role to play in the licence for principles that would have given private copyright holders around the world a legal weapon with which to deal with spyware problems.

I think those problems are terribly severe. I expect most governments to get it wrong and keep it wrong, because their incentives are misaligned: the victims of spyware are individual citizens; the beneficiaries of spyware are wealthy public and private parties with enormous benefits to gain from the continuance of low security; as Microsoft fails, and the primary provider of low network security around the world begins to fail; the parties which have benefitted from ten years of low network security, and boy have they benefitted!, will begin to look for other ways to insure low network security.

The monopoly was an extraordinarily good way to achieve low network security, because the only party who had to take the blame, was the party to whom there was no alternative, and which was perfectly happy to have its non-existent reputation for integrity blacken still further. Once that is no longer true, the problem of how to achieve high network security given many commercial parties' desire for weak network security will grow more intense.

I thought it was desireable to afford individual developers the opportunity as private attorneys general around the world to use copyright law to inhibit bad practices built into their code by subsequent modifiers. I now conclude that that's a provision for which, by large, the community does not see it in that way, and it is not hard to remove it, given the absence of people's interest in it. Later I believe they will wish we had done that, probably in the same intensity that they would wish that we had solved the patent more coercively in 1991, and may very well be glad that we have solved the DRM problems for them to the extent that we are going to solve them this year. But we will leave that problem to be dealt with by the future, because it's clear that the readers of the licence in the present don't see its value, and do see significant drawbacks, so I believe that it will come out.

[56m 40s]

Q6: Will we be able to use the same methods used to make the LGPL to make the new GPL compatible with the currently incompatible licences like the MPL, the Mozilla Public License, or the...

[57m 00s]

You can always use an additional permission, to combine GPL'd code with any other licence that you want. So in that sense, yes, you can use additional permissions to make trivial compatibility between GPL and other licences. Please remember that permissions are removable: downstream from you, parties may decide to remove your trivial compatibility provision, thus returning the code under the licence to a pure GPL state. Remember that existing LGPL contains a provision that says all code licensed under LGPL can always be relicensed under GPL. So in turning LGPL into an additional permission, we are not changing any behaviour by making it a removable permission on top of GPL. If you used a similar approach to make compatibility with another licence, and admitted the removability, you would be fine.

[58m 05s]

Q7: How are you thinking about changing something in the title of the section, I think it's 9, "not a contract", because that that's a bit incompatible with the laws in some place, like in Brazil - I'm from Brazil.

[58m 23s]

I firmly disagree with that position, but nonetheless we will do something to meet these needs. I ultimately regard these believes as narrow-minded and foolish, it belongs to half a dozen law professors around the world, each of whom should check his cards again, and it's all "he". They are people with gray hair and old minds. They're not very old minds, because if in each of their legal systems they went to their old legal dictionaries and looked at what the word licence means, or if they got real Roman about it and went and looked in the Institutes of Justinian to find out what licence means, they would discover that a licence is a unilateral permission, not an obligation, and so what happens is that these minds that say these thing, they're stuck in a little space, a thousand years after Justinian and before the Second World War.

Nonetheless, there's a view, OK? And so that view, which has been hardly repeated, deservers to be regarded with respect, though I haven't given it any respect in my personal comments, because I'm entitled to a personal opinion. As a working drafting lawyer, I expect that section 9 will bear, when it is released, a title something like "acceptance not required to receive and run copies", which is a statement of fact, accurate and clear, and which reduces the reliance upon a legal characterisation, increasing reliance on a factual consideration, which is in general the course we have taken with respect to de-nationalising the licence.

So much as I believe as one working lawyer that it is a foolish legal position, I recognise that it is a straightforward request that can be served and that should be dealt with, and we will deal with it.

[60m 40s]

Q8: I have two questions. First, do you have anything to say about the so called "ASP loophole", or "web loophole" because in the morning RIchard said that he saw the possibility to adopt some clauses from the Affero GPL to fix this problem, but it may be too drastic for GPLv3, but I think this loophole can be very dangerous. Some day, our friend Microsoft will be gone, but we might have the next enemy, because now we have the so called "Web 2.0 companies" whose business model is basically built on web services. Would you answer this question first?

[61m 40s]

Knowing, as I do, that you may very well wind up not getting your second question in, but that's all right. Look, here's the situation as I think we know see it: there are two ethical points of view about what should happen when you take somebody else's code which they wrote to interact with users over a network, modify it yourself, and go into business competing with them to serve the same people using your modified version of their code: one ethical position is "you should be required to give your changes back to the world: you're competing with others in the public provision of services based on that code, and you should return to commons your modifications, the same as if you were selling or distributing that code", that's one ethical point of view.

The other ethical point of view is that this is a uniquely distinctive situation in which any attempt to interfere with the right to keep those subsequent modifications private, interferes with the right of private execution and private modification, rights that are not to be tampered with. Those ethical points of view are both very strong. THere is no benefit whatever to the licence maker, or to the free software world in general, in choosing between those ethical theories for everybody else.

If we came to you and we said "we thought about these hard ethical issue; there are absolutely good and absolutely clear arguments on both sides, they are of equal weight and importance, this decision will have billions of dollars effect in the world, we've decided it this way, A way or B" people would rightly say "what had you in your world, that justified you with making that decision for everybody else".

[63m 50s]

We will not make that decision for everybody. Nor will we make parties who want to adopt either ethical theory forego all use of existing GPL code. Imagine for a moment that we took out of section 7 the portion that relates to the so called Affero GPL clause, we dropped it from the licence: now anybody who wanted to, could make a new licence containing that provision, but it would be GPL-incompatible, and they could not use any of the world's existing GPL'd programs as a basis for building on.

That would inevitably harm innovation, because it would deprive people of all this shareable software. On the other hand, if we decided that every GPL included the Affero clause, there would, as you point out, be an awful lot of businesses in the world that would lose their right to use GPL'd code, because for reasons that they consider good and sufficient for themselves, they do not wish to make their private modifications, which convey business advantages to them in public facing systems, open to everybody else; they want to keep some source code that they have modified, even though they are using that source code to relate to the public.

Very well, we say; we would simply allow an additional requirement to be placed on some code; if you build an Affero GPL model system, that is one which implements services over a network, deals with the public over that network, and in which any competitor who wants to use your code to build similar services must release their modifications, the best way of doing that is to build your whole application under GPL, take the part which does remote network services, which may be just one among many modes or models of interaction for that program, put that piece under Affero GPL, or add the requirement, is how we would say it, and add to that program a feature which allows the user to download the server side source.

Once you've taken all those steps, your application can be made of any GPL'd parts yours or other people's; the particular part that you wrote to communicate with the user over the network bears some additional requirements. Anybody who wants to take the whole rest of your application can do so and use it in any way they want, including private modifications and public uses.

And, if they don't like the part that you wrote being under Affero GPL, they can rewrite it from scratch, thus minimising the amount of re-implementation necessary, in order to take your application, which was Affero-lised, and de-Affero-lise it so it can be used for private services. After lengthy analysis we have concluded that this results in the optimum level of overall innovation and the minimum level of required re-implementation, so that every developer can get the result that every developer, or that developer's business, wants to get. I do not think that there is another solution available which reduces the amount of required re-implementation or which increases the net total of innovation, therefore I think this is the right system to apply.

[67m 30s]

Q8b: But we take a very firm attitude toward DRM, right? And I think, from a user's view, if we can use the software, I mean the service, and we can get the results, but we can't see the source code, it should be called proprietary, and it may infringe our the third freedom, to help the community. So, maybe I'm too dogmatic but...

[68m 10s]

I don't understand how you are loosing any rights. If you write the thing yourself, nobody else's code is involved and you can make all the rules that you want. If you use other people's GPL'd code to make an application with, how does it harm your rights to say you have to respect your users at the same level that they respected you as a user when they gave you the code.

We're not in the business of preventing anybody from doing anything that he wants with something he writes himself. The question is: if you get stuff from somebody else which constitutes an entire application for public facing services, and if he has written his application so that he wants mods released back into commons, why should you think that you have a right to deprive him of that, and still use his code?

My notion would be: if you want to re-implement the part of his code which uses network services and leave the whole rest of it in place, you get a great deal of benefit, you get to use 90% of his code; he gets a benefit: you are required to respect the user's rights for that 90% to the same extent that he respects the user's rights. If he respects the user's rights a little more in the remaining 10% than you wanted to respect them, you have a choice: swallow hard and take his approach to your users, or throw away that piece of his code and re-implement. That's the minimum size re-implementation for you to have the same absolute freedom that you want to have; I don't think you're disadvantaged, I think you are advantaged to the extent of the continued shareability of the code between two fundamentally ethical positions.

[70m 00s]

Q9: If it's OK, I'd like to ask a second question, about the translation. You said that translation are second best, and I think I agree, more or less, but I'm from Japan, where most people, even those who can speak English prefer to read the Japanese translation of the licences. I understand the idea of official translations may be very dangerous, but I think that it would be a good idea to make quasi-official translation, which means, it's reviewed by several local lawyers, Japanese translations by Japanese lawyers, Spanish translations by Spanish lawyers, and at least FSF acknowledge that this review has happened.

[71m 00s]

I've two things to say about that. The first is that, as Mr. Stallman indicated today, the question of translation is open for discussion, and I am listening to everybody about it. The second thing I want to say is that I made a trip to Tokyo before the beginning of the GPLv3 process, to meet with members of "Meek" and "Makey", and to meet with lawyers from all the major industrial organisations using GPL in Japan, they were all very gracious with their time.

I said: we're going to make GPLv3, we badly need Japanese lawyers who understand this business and its concerns to help us think about every single word in the GPL from a Japanese legal perspective; I want to work with the best Japanese lawyers you all identify to make this a better licence; they all said "Thank you very much, we'd be delighted to do that", and they did nothing. And the reason they did nothing was they did not want to expose their lawyers for non-privileged conversation on these questions.

The result is that 12 months after that meeting, which occurred in June of 2005 in Japan, I have so far had zero meaningful legal assistance from the lawyers working for large Japanese industry. If it's important, it's important. If it's important later, it's important now. If it would be good to have them make an official translation later, then it would be good to have the same lawyers involved right now helping us to make a better licence.

[72m 30s]

Every single large industrial organisation in Japan has highly talented lawyers thinking hard about these questions, and they speak English quite well, as I know from having done business with them for years. The problem we are having is not a language problem, the problem we are having is not a willingness to listen problem, the problem we are having is not a "we don't care about Japan" problem, the problem we are having is the parties have decided to wait and keep silent too long.

I wish they wouldn't. I think the most useful contribution you could make would be to put heat on the organisations to bring their people in and start negotiating. It's now more than ready to start happening, we're halfway through the process; if my good friends in Japan wait until the eleventh hour plus thirty minutes to come and tell us what they think and what they want, they're going to be disappointed; not because I'm not going to listen, but because the train will have left the station. Please help them to get on board.

Q9b: That's what I'm doing now, I actually agree.

I'm responding to the suggestion that the way we should let them participate now is by letting them translate it when it's over. And I have said "Gee, that's too bad. That's not a good way. Let's do it now!".

[74m 20s]

Q10: You opened and closed your talk, talking about Microsoft capitulation. As someone who has constantly to deal with Microsoft meddling with governments and blocking initiatives by sheer amounts of money power, and stuff like that, I'm more than keen to believe that it is such, and it would be really, a great help for us not have to constantly fight that; however, from being exposed to that kind of thing too often, I'm not very inclined to be so optimistic as to it going to stop anytime soon. What should we be looking for in the near future as further indications of this capitulation actually being a fact?

[75m 25s]

Well, there used to be a Microsoft executive, who was the executive vice president for platform technology, whose primary task was to "beat Linux", as they described it. His name is Martin Taylor and everywhere he went, Martin Taylor was a very important part of the landscape of pressure and interaction that you have described.

After the second great [inaudible] when Windows would only be a little late, Mr Taylor was moved to work on Windows Live, of which he became the dominant executive.

Earlier this week, suddenly, without any pre-announcement of any kind Microsoft fired him. Mr. Taylor has worked at Microsoft since the age of 23, he's now 36; he was personally a pet of Mr. Ballmer's; he has a great deal of Microsoft stock and no job; I don't know whether you would consider that a straw in the wind or not.

I would say this: among the reasons I wanted to do this work now, one of the reasons we identified when we started to pick this year to do GPLv3 was that I believed that Microsoft would be busy doing something else. Their position in our process, their power to disrupt, was a significant concern to me from the beginning, when a company is trying to ship a major product of its own, of a kind that demands months and months of careful marketing and preparation, PRs as well as merchandising, it is not likely that they will spend a lot of money saying aggressive bad things about a third party.

It's not known to be a good way to get your own product to market, to spend a lot of time bad-mouthing other people. So I thought this would be the best possible year to do this work, simply because Longhorn, now Vista, had to be released this year.

[77m 25s]

I now believe that it runs a little deeper than that; the charm offensive Microsoft is currently conducting, which is palpable to any of us who work in this part of the world, the charm offensive Microsoft is conducting is now edgier than it used to be for them, there are fewer alternate options.

Of course I don't want anybody to come off alert; I'm a lawyer around here, my job is anticipating war and being ready to fight it whenever it turns up. We have by no means stopped worrying about aggression. When Samba 4 is released, Microsoft will have the best reason for aggression it has had in a long time, because it will be competing against something that doesn't just ruin the server business for them, it eliminates the server business from them.

The consequences are not going to be small for a company which already has nothing to offer on the server side, and which has never successfully sold the client side of anything to global enterprise without a server side to go along. That's why the current situation is so dire in business terms for Microsoft; that means that they will be, when they act aggressively, if they act aggressively, acting aggressively from much greater weakness that they had in the past.

Moreover, thanks to Carlo and others in this room and around this room, Microsoft is far more constrained by the European Commission anti-trust proceedings than Microsoft was ever restrained by any anti-trust proceedings anywhere else in the World. For the first time, the company faces the possibility of significant alterations to its behaviour mandated within the lifetime of one of its products.

And Vista, a product they are not going to be able to sell to knowledgeable consumers, is also in the pipeline at just the moment when it can be decisively affected by regulatory orders from the European Commission, and Carlo and our colleagues in that litigation have already shown that they know exactly how to testify, and exactly what to ask for, and exactly to hold on to the ankle without letting go.

[79m 35s]

I'm not here to tell you that this is going to be easy or simple, or that the US embassy is never going to be asked again to do a favour for Microsoft in some country in the World where the US embassy's word is terribly important. All of those things may very well happen. Our general ability to return fire is superior to at anytime in the past; our ability to do real strategic damage to their earning capacity is stronger than at anytime in the past; our allies more firmly believe in the possibility of beating them than at anytime in the past, and in Google they face an adversary of whom they are terrified.

Google is from our point of view just a large user of our stuff, but from their point of view Google is the business that terrifies them; and therefore, they are now dividing their anxiety again for the first time in years. After Mr Gates figured out that we were the problem, we were for a long period of time his primary anxiety; now, it doesn't matter what makes him anxious in the middle of the night, he's just a philanthropist. As for Mr. Ballmer, what makes him anxious in the middle of the night? Lots of things make Mr. Ballmer anxious in the middle of the night; we're going to keep pumping up his blood pressure until maximum is reached, very soon.

[81m 10s]

Q11: This is a very basic legal question, I am not a lawyer...

In this country I'm not a lawyer either...

[laughter]

Q11b: I'm an eager reader of Debian Legal, which may be a bad thing, but I'd like you to comment, as regards to the GPLv3, on what's called in Debian "the dictator test", or the tests that regard something that is not distribution or dissemination, as would be for instance the patent retaliation clauses. I don't completely understand about how these things can be free, which are really restrictions of freedoms that the person would normally have and do not affect pure redistribution; one is patent retaliation, the freedom to sue is one that one has, supposedly, and the other one is the Affero type of thing; Affero affects use, or liberty/freedom 0, not distribution or dissemination which would be freedom 1, as far as I understand. Maybe you could comment on that and why the GPLv3 takes the view that it takes on those two issues.

[82m 20s]

Look, there are two aspects of that question, one of which i think is, as you say, the Debian Legal way of life. I try not to interact with the Debian Legal way of life, but having just said I'm not a lawyer in this country, I suppose there is a reason that in this country I might interact with the Debian Legal way of being a little more; I'll try and pretend that I don't know anything about law anywhere in the World and then I'll try and think Debian Legal's way about the stuff.

It generally turns out, as I know from having spent almost a quarter of a century now as a lawyer for hackers, that when hackers pretend to be lawyers, there are certain predictable formulations that they come to; they assume a degree of consistence in legal rules that is not achievable; this is a primary problem which occurs particularly in US focused conversation, such is that in Debian Legal, where the libertarian demand for intellectual consistency, and the hacker belief that laws are form of code that are executed without errors or ambiguities, joins together to create a particular frame of analysis for legal questions.

It doesn't work very well for me as a lawyer, I think it doesn't work very well for lawyers elsewhere in the World, because the one thing which lawyers around the world all share is an awareness of the squishiness of law, it is by no means the hard arthropod carapace for internal soft organs that non-lawyers have a tendency to assume it is.

[84:00]

With that in mind, the problem of the patent retaliation clauses is not a problem in the GPL, for Debian or anybody else. If the Debian legal group had decided that the Apache Software License was a non-free licence because it had a defensive patent retaliation clause in it some years ago, maybe the Apache tribe would have decided not to put such a clause in its licence. We are merely saying we are prepared to coexist with it, and since the Apache licence is already a free licence, from Debian's point of you, I fail to see how coexisting with what's already free could be a freedom problem.

[84m 35s]

So far as the Affero GPL clause is concerned, remember that what it says is: "if you get a GPL'd program, that has a facility in it, giving server side source code to the user, that's not a removable facility". That's the same statement that the GPL makes about interactive display of copyright notices; lots of licences in the world graded free say "if the program displays copyright notice when it starts running in the form you get it, then you have to leave that there", and the reason is for security of attribution, right?

So here is another non removable element in a free program, placed there out of the desire to protect people's freedom. I don't know how the Debian Legal World will think about that, I know how the Free Software Foundation will think about that; the Free Software Foundation would say: the very fundament of copyleft is the desire to limit freedom minimally in order to protect freedom in the long run.

The question to be asked within the Free Software Foundation's ethics of law is not "Is this a restriction on freedom?", which is what I think I heard the Debian Legal analysis is doing, the question is: "If this is a restriction of freedom, is this the narrowest possible restriction to protect freedom in the long run, or is there a narrower way to accomplish the same protection of freedom, or is it better to allow freedom to be defeated in the long run than to impose this restriction now?".

But it is always, I think, in Free Software Foundation thought, it is certainly characteristic of Mr. Stallman's thought throughout his career, it is characteristic, I think, to be minimalist about restriction, not negative about restriction.

The position being taken is "to have a community requires some minimal restrictions to preserve community". We tend to see the other side's view of that question "there ought to be no restrictions except those which are inevitable" as too libertarian to be firmly communal. If you want to keep a community in being you may have to say "clean up after your dog on the sidewalk". Yeah, it's a restriction on use, but it's an alternative to having a lot of dog shit on the sidewalk.

[87m 10s]

Q12: Just a small speculation. You said that the licence is a unilateral permission. Why is that so hard for most law professors to understand? I agree with it, but I just don't understand why very few would like to argue in that direction.

[87m 25s]

Because in the course of the 20th century, post Edison, the copyright licence became a device for getting more out of the customer than copyright allowed. Once the Edisonian model of culture and distribution had come to be second nature, very few publishing parties wanted only what copyright gave them, for multimedia or technically sophisticated works, they wanted more than copyright could give them, so they entered into contracts with their customers, one part of which was the licence, and another part of which was additional obligations that they wanted users and customers to have

So over the course of decades, a copyright licence ceased to be only a licence, and became very largely a contract for customers' obligations, because the customer's obligations were a big part of what the party writing the licence, which wasn't the customer, wanted to see in there.

We came along and said "We don't even want all of what copyright gives us, let alone do we want anything else", so we can use a pure licence, because most of what we want to do is to give away copyright law, and no part of what we want to do is to take additional obligations from people that aren't within the scope of copyright law.

Therefore, for the first time in decades, somebody was building a business model around a pure licence, and a whole bunch of lawyers who grew up in the world after Edison and before Stallman said "we've never seen one of those! Every licence must be a contract, because every licence involves obligations by the licensee", and we said "That's new! That's novel! That's crap!".

Users have rights: you shouldn't take people's rights away from them using contracts, where what you're giving them is just a program to run. It's in the nature of the criticism we were making of the legal institutions being used by others, that we were calling attention to a thing which had happened barely consciously, and which is still a kind of unreleased restriction in some people's minds.

Alright, thank you very much, I appreciate it.

[Q&A ends, applause]

Further information