Sean Daly of Groklaw, who is also a member of the FSFE, interviewed Free Software Foundation Europe President Georg Greve at the end of day four of the hearings before the European Court of Justice in Luxembourg regarding Microsoft. Here is the interview as Ogg and here it is as MP3. He also provides a transcript. Greve explains some of the issues that have been raised during the hearings, such as interoperability, Microsoft's patent claims as revealed in court, the documentation problem, and why reverse engineering is a game of perpetual catch-up, and gives his impression of how the hearings had been going to that point.
Interview transcript: Georg C. F. Greve, President of the Free Software Foundation Europe
April 27, 2006 11:00PM interview, near European Court of Justice, Luxembourg
Interviewer: Sean Daly
Q: OK, I'm sitting here with Georg Greve of the Free Software Foundation Europe, he's in fact the president of the Free Software Foundation Europe and he's had quite a long day at the hearings in Luxembourg for the Microsoft appeal and I should say he's had a long week because prior to that he was in Brazil [for the Fórum Internacional Software Livre]. Now, I have prepared a few questions and why don't we get right into it? Now, first of all, what can you tell us, how has it gone today? How has it gone these past few days? What's your general impression about how it's gone?
GG: My general impression is more or less positive. There's always things that you wish had gone better. There's always things, you know, where you think, you know, that did not necessarily, you know, give the right impression. You always kind of try to wonder, you know, to see how the judges are reacting and wonder how they take it, you know, what impression do they get from the different statements. Overall, I think it looks good. I'm fairly certain that we do not have to be pessimistic about the outcome of this case so far. We have not screwed up royally; all the sides on the side of the European Commission did a good job; one or two things may have been improved possibly, but I think overall we had a good day.
Q: So you could say you feel that your voice has been heard?
GG: Yes, absolutely. The voice of the parties on the side of the Commission, including those of the Free Software community, have definitely been heard.
Q: Now, these past couple days the hearings have been concerning interoperability. Perhaps you could tell us in a simple way, I mean, interoperability, what does it mean -- it's a long word -- when we're talking about interoperability? Of a Windows server and a GNU/Linux client? Or are we talking about a GNU/Linux server and a Windows client? What do we mean by that?
GG: Well, first of all, interoperability is a rather difficult subject in many ways to begin with because there is no thing as an absolute interoperability. There's different forms of interoperability and they gradually differ. What... if you want to understand what interoperability is
you can think of languages maybe. It is, essentially, that two machines, if they are to be interoperable, should be speaking the same kind of language and agree on the meaning of that language. Essentially like two human beings, if we say, "this is a table", then we both associate this object with that word "table" because we have a prior agreement on what kind of object has which word and so on.
With computers, it's not much different in principle and the interoperability depends much on which kind of language the different machines speak. So, a programmer, in this case the programmers of Microsoft, can choose which language to implement in their machines and how fussy they are about, you know, adherence to that language. So, if I for instance said "this is 'une table'", you might still be understanding that I meant to say "a table", because your fussiness level is lower because you want to understand me. You could also prefer to be extremely fussy and only accept that designation of the table as a table if I say that incredibly correctly to the standards that you expect, which is where the problem starts, because I have to know which standards you expect. And that kind of information is exactly the kind of information that Microsoft does not give its competitors. So the competitors do not know what kind of language the Microsoft machines speak and how exactly they expect to be addressed, and you know, what kind of objects mean what kind of things. You can't really know that; you can watch a conversation between two Microsoft machines and try to figure that out, which is protocol analysis, also called reverse engineering, but that is an incredibly ineffective way of doing it and it takes a long, long time.
Q: Now, I understood from news reports that Microsoft had previously been better about interoperability up to, I guess, 'til 1998 or so and then apparently they decided to start slacking off and making it more difficult. Can you tell us anything about that?
GG: Well, yeah. I mean, Microsoft originally had big interest in interoperability because they played no role in that server market. The workgroup server market was one that, you know, was essentially considering Microsoft to be more or less irrelevant. People interoperated with Microsoft desktops when they started, you know, to have network functionality, but the [Microsoft] workgroup servers didn't really make much headway. Novell was rather strong in that area and so were a couple of others. So, Microsoft had a big interest in having a common kind of standard to talk, you know, to work with servers for the clients, and indeed started a standardization effort. There is the effort around the Common Internet File System, the CIFS, including the CIFS Conference which Microsoft, I believe, held for the first time. They were the initiator of that effort. And, well, so in the beginning, they actually did have a lot of interest in this and even released that information -- which is the same kind of information that we ask for now -- they released this information to, for instance, the Samba Team which has been implementing it into the Samba software. So, yes, in the beginning Microsoft was cooperating nicely with all the other vendors, as all the other vendors are still cooperating nicely, only Microsoft now locks other people out with arbitrary changes to their protocols that you cannot easily reverse-engineer, where you always have to play catch-up and in principle, so disadvantaged that you have no way of building a product or any kind of software that would be able to interoperate with these Microsoft machines in a way that would at least come close to what Microsoft can do.
Q: OK, let's talk for a moment about Microsoft's compliance in that area. I mean, you know, they love to mention about forty times about how they created a 12,000 page document that can help just about anybody interoperate with their servers. I mean, not having inspected it myself, isn't 12,000 pages enough?
GG: Well, I would say it's too much! I mean, there's a lot of things that you could be worried about in regard to these 12,000 pages. It begins with the fact that Microsoft -- according to its own claims they have made in court these days -- had to spend a lot of manpower on the project of documenting its software, including hiring retired engineers who had left the companies so they would tell Microsoft what was happening in their software. Now, the total of that effort -- allegedly, I mean, I can't check this -- was 35,000 person-hours, to document what they should already know. Doesn't Microsoft know what is happening inside the machines that are on 90 to 95% of the desktops on this planet? That is worrisome, because documentation of specifications and interfaces is part of good design and good interfaces are a necessary part to have stable and secure systems. The second part that could be worrisome about the documentation is the fact that, you know, so many pages mean that that is a huge heap of information
which makes it extremely difficult to parse or in any way use that information to actually do something with it, meaning writing software that is interoperable. So, what did Microsoft do, according to their own explanations in court? They said that this was so unwieldly that they put it into a digital form which used a specific kind of sorting method, a specific kind of, you know, tree algorithm that they claim to have software-patented
so without a license to that software patent you cannot even parse that documentation and if you do, they made it rather clear that they would reserve the right to sue you.
Q: So you're telling me that this is not like a printed binder that I can open up to page 47, but it's some software and I need to get some license just to access the content of the software [documentation]?
GG: Essentially, that's what it boils down to. I mean, if I understood the statements of the Microsoft lawyer correctly, that is indeed what they said in court. And they were relatively explicit on this, so I don't think that I really misunderstood them there.
Q: Well, let's talk about intellectual property for a moment, because Microsoft keeps talking about "IP", and "intellectual property", and it's very, very important, and yet, I find that "intellectual property" is a bit of a catchall term and we could talk about copyrights, you could talk about trademarks or trade secrets, or patents... When Microsoft is talking about "intellectual property," what is it they're really talking about, in your opinion?
GG: Well, to begin with, I have severe problems with the term "intellectual property", because it does lump together different areas of law that actually are rather different. Copyright is a different area from patents, is a different area from trademarks, which altogether are a different area from geographical indicators. And all of these are often lumped together under this heading, conveniently so for people like the lawyers of Microsoft because they can use that expression, that term, to confuse the people about what they actually talk about. You know, it always sounds like "oh, you're taking away our property". In fact, the word "expropriate" has been used repeatedly in court today. And that is only because the term "property" suggests this was some kind of, you know, material property, whereas indeed these are monopolies, you know, limited monopolies granted by society for the benefit of society, in theory. That is what they should serve -- they should serve the benefit of society. Which means that if they stop to serve that benefit, they lose their justification. These are monopolies that we grant as a society to individuals on that basis. So if that basis goes away, the justification goes away for these monopolies to exist. Microsoft has tried to make claims to copyright in the beginning, saying essentially, "Oh, these people just want to see our source code. That is a copyright problem. You are infringing our copyright. You want to take away our copyrights", and so on, but they couldn't really maintain that position because it was extremely clear that source code was never the subject matter discussed. No one wanted to see their source code, essentially.
Q: This was what you meant by the "poisoned honeypot"?
GG: Well, sort of. The true poison in that was that they had a WSPP [Workgroup Server Protocol Program] kind of specification licensing agreement that they had proposed, and if you had agreed to those terms, you would have agreed that you would not use what you learned to write Free Software. So, a Free Software developer that would have agreed to those licensing terms would have found themselves unable to work on something in this area implementing Free Software. So, that, you know, is a kind of tainting in a legal sense that makes it impossible for people to do what they really want to do.
Q: Let's talk about licensing again. I mean, the impression that I had -- I looked at some of the liensing terms on the Microsoft site and they really seemed stacked against independent programmers who are working collaboratively on the Internet. They seem to be designed for companies. And I think it was Simon Phipps of Sun who talked about the "Non-Entity Problem". He said, you know, if you work for an entity, a big company, this company can afford a $50,000 license spread across all their programmers. But if you're one very talented person creating software collaboratively on the Internet, with a wiki, with modern tools, there's a real problem. Do you feel that Microsoft has specifically tried to exclude Free Software with their licensing?
GG: Yes, but not with the -- necessarily -- with the financial component of that. The most important implementation that is interoperable to a certain degree with Microsoft software is the Samba software. And, indeed, we have been working extremely closely with the Samba team throughout the entire case and the investigation that preceeded it. Now, the Samba Team could have fought to pay a one-off licensing fee to get access to that information. That would not be a big problem. We could find that money. And ultimately, it would be better for us to find that money with some of the companies that support Samba, that have Samba developers working for themselves, and they would pay that money. There's no question about that. That could work for Samba as long as it is not a per-seat licensing fee, because that can obviously not work for Free Software. The amount of potential users is, you know, well, the population of this planet and so you would almost have to take that as the base amount of users to, you know, calculate your per-seat licensing in a way which, you know, is unaffordable even if you have a very low amount. I mean, you cannot control this -- in fact, you should not control this because it is Free Software -- you should be able to use it for any purpose, you should be able to distribute it, you should be able to modify it, you know? It is necessary that Free Software gives you these freedoms, otherwise it is not Free Software any more. And therefore, you know, it's not acceptable.
Q: Well, I believe it was Ian Forrester, one of Microsoft's lawyers, who said, "Well, our competitors -- Microsoft's competitors -- they just want a free ride". So, if I wanted to play devil's advocate and say, you know, "I have spent billions paying for researchers to work for years to create this fabulous software. Why should I just give it away?"
GG: Well, first of all I would tell you that you're not supposed to give your software away, because we don't want your software. We want to write software that can talk to it, but we don't want your software. That is number 1. Number 2 is, pretty much all of what you have in your communication protocols has been taken from open and documented standards: Kerberos, LDAP, and a couple of others. All of these are in there. What you have done to them is you have added small little extra things that you did not document. You have changed the protocols in small but fundamental ways that make it impossible for those who implement these protocols in a clean way to actually talk to you. And if that takes you hundreds of person-hours to change, you know, small bits and pieces in that protocol implementation, then I think that you have either spent that time playing Tetris, or done whatever else, but ultimately, you know, these protocols are not secret because they are valuable, they are valuable because they are secret.
Q: OK. Now, I've understood from press reports that the European Commission will be prepared to make a judgement concerning fining Microsoft for noncompliance by mid-June, I believe, whereas I understand also there won't be a result of this decision, on this appeal, for many months and possibly a year. So what do you think is going to be the next step? What happens next?
GG: Well, that is very hard to predict. It depends on a lot of factors, many of which are far outside my control and even far outside my range of information. Therefore, it is very difficult for me to predict exactly what is going to happen. It depends a lot on Microsoft as well, and I cannot predict what they are going to do. I mean, I hope, obviously, that the Commission will maintain its firm stance and will apply those fines because, well, Microsoft has worked extremely hard to deserve those. So I think they should get what they deserve.
Q: Don't you think -- I mean, we often speak of the massive cash reserve that they've put aside. Don't you think they could laugh it off, pay the fines?
GG: Possibly. But one should not underestimate the damage to the image that comes with being a convict essentially, you know? You are guilty of, you were found guilty of abusing your monopoly and you are paying a daily fine. That does not look too good in the press. And not too good to your customers.
Q: Now, on a different subject, I understand you were present for at least one of the days dedicated to the Windows Media Player tying. Are you aware if there was any discussion concerning privacy issues around Windows Media Player? My understanding is that the more recent versions of Windows Media Player in effect give Microsoft an opportunity to track what it is you're doing.
GG: I am not aware that such discussions have taken place. I do believe that they should have taken place, as I believe that a couple of other discussions should definitely have taken place here which also were completely ignored ...
Q: Like what, for instance?
GG: Well, to begin with, obviously the software patent issue that has arisen here because, you know, Microsoft has made it so clear today that they consider themselves, at least, to be holding around 46 software patents on various things surrounding their Active Directory service and that they would enforce those patents if need be -- or rather, that they consider if the court orders them to surrender that information allowing interoperability to have been expropriated or compulsory-licensed for those patents, well they don't really like that at all. They have announced that ...well, essentially made it very clear that they would do whatever they can do to prevent people from cooperation and competition through these software patents. And you know, it is difficult so far for me to listen to this, everything in me screams that someone should put up the quote from Bill Gates in which he said that "if people had taken out patents at the day when today's inventions were made, then actually the computer industry would be at a total standstill today" -- that was paraphrased, but you can look up the quote, I'm sure you'll find it. That's also once or twice in my blog, actually. And really, everything just screamed that because it was so clear written all over the courtroom, yet at the same time I do understand obviously that the judges cannot make the politics, because they have to be independent and implement what is there in law. But still, it is sometimes hard to listen to it.
Q: Now, I understand that Andrew Tridgell of Samba was in court today. Can you tell us anything about what he said?
GG: Oh yeah, Andrew Tridgell was there and he made a fantastic impression too. Start off with that. We were extremely happy to have Andrew Tridgell over to speak on behalf of the Free Software Foundation Europe and the Samba Team which you know have been working together throughout the entire case. We had Jeremy Allison over for the interim measures hearing and so this time it was, you know, the founder of the Samba project. Andrew Tridgell is, you know, uniquely skilled to talk about these issues; also, he is a very good speaker and, you know, intimately knowledgeable in the whole area, so could easily rebut anything that Microsoft's experts were bringing up in terms of fog and smoke grenades and put a lot of clarity, I believe, into the case. He explained how Microsoft's "Blue Bubble" which they had invented for the sake of the case to claim that there could never be interoperability between directory service, that this "Blue Bubble" actually did not exist or as he said, it is a "blue bubble of secrecy". He also explained how the Samba Team had been struggling over the years through protocol analysis to build something that was fully interoperable with Microsoft, but it's now more than ten years behind essentially. The problem is that protocol analysis ultimately is a very clumsy way of doing this and even though the Samba Team is without question the group of people that are, you know, uniquely skilled to do this job and among the best experts in the world to do this kind of job, they've been working on this six years and they are still years behind.
Q: Well, you know, that reminds me of a post that Andrew Tridgell put up a couple of years ago where he said it's like figuring out what the French waiter says by just sitting down in the café and every time people ask for something and he brings bread, that you figure out that this is what the word "bread" is.
GG: Yeah, kind of like that. Indeed, it would more be like -- sometimes even like the waiters are changing, from a French to a Spanish one, so you have to relearn the word "bread" every single time. Because Microsoft keeps changing these things and indeed keeps also adding fussiness to their clients in terms of how exactly they do expect certain responses to be. So, through that client monopoly, they were able to dominate the server monopoly which is what the entire case is about, and I believe Andrew Tridgell was very good at explaining this, and also explaining to the Court in particular that this is not about "free riding" as Microsoft has alleged among other things, but rather about allowing competition and innovation.
He showed a very small palm-sized box which is a router or actually not a router but a router-sized device that is a -- well, you can think of it as a smart disk and the "smart" in it is Samba. You can plug USB hard disks into this device and put it in your network and embedded on that device runs Samba and serves the disks on the network. And that small box could possibly become an Active Directory server for several hundred people and could do all that work in this palm-sized box for which Microsoft right now requires a full PC. I mean, Samba has hardware requirements that are more than ten times lower than those of the Microsoft implementation. And he was able to explain this, and indeed I think you should probably talk to him to hear more about the details of this because he has a very clear way of saying this which is far superior to my way of explaining it. Plus I am sure he will not be tired when he gives you those answers. But ultimately I think that made a huge impression on the Court because it showed that this is not about "free riding." This is about coming back to the industry standard of cooperation and interoperation. This is coming back to allowing competitors to actually compete with you. This is coming back to allowing innovation. That is what this case is ultimately about.
Q: OK. Thank you very much.