[LWN Logo]
[LWN Feature]
Weekly Edition
Daily updates
Events Calendar
Book reviews
Penguin Gallery

About LWN.net

December 10, 2000
Maya Tamiya

Eric Raymond Interview

This interview took place on November 30 at Linux Conference 2000 Fall, which was held from November 29 to December 1 at Kyoto International Conference Hall, Kyoto, Japan. The event was sponsored by Japan Linux Association and held with the Perl/Ruby Conference by O'Reilly Japan. We would like to thank Eric Raymond, Takaaki Higuchi, Tomoko Yoshida, Mayuko Yoshida, and people at JLA who helped realize this interview -- Thank you very much!

We talked about many things including the secret of success, licenses, patent issues and so on.

Eric

CL: You contribute to many projects including big ones like Linux and Python, you also have projects you manage, you're one of the board members of VA Linux, you are often on the road to give talks and you still take time to write open letters. How do you manage to do all of those things?

Eric

ESR: ...I just do them... This is my life. This is what I do... I find the time to do the things I'm interested in, by refusing things to do I'm not interested in.

Most people's lives are taken up with a great many trivial things that they don't really care about, but which they feel they have to do. I just don't do that.

I don't have many physical things, which means I don't have to spend a lot of time taking care of and worrying about physical things.

CL: But you have a cat.

ESR: That's true. The cat officially belongs to my wife, though. She takes care of it.

Other people have lots of clothes, I don't. Other people have cars, I don't. In America, a lot of people worry about what their lawns look like and spend a lot of time mowing their lawns, I don't do that. I hire people to do that for me :-)

A lot of people think they have to have jobs. I gave up having a job fifteen years ago, and never regretted it :-)

CL: But then you have to start worrying about other things such as money?

Eric

ESR: That's true. But it's been true, at least for me, that every time I have concentrated on what I really loved, every time I have done what I really cared about, the money has come. Somehow.

CL: What's the secret?

ESR: It's just concentration. Focus. Keeping your mind on what you really want and refusing to be distracted by other things.

I left my last job in 1985. And I started making my living by doing contract programming, writing books, and doing technical review work, and concentrating on things that I really cared about. When you concentrate on things you really love and care about, you do them really well. There is usually money waiting for anyone who can do something really well.

You see, that's why concentrating on what you really love is important. Because if you give a talent or skill total concentration, chances are you will get good at it to attract money with it.

Maybe I was lucky, but luck only helps when you put yourself in a position to benefit from it. In English we say "Chance favors the prepared mind". Perhaps I've been lucky, the point is that I created the conditions under which luck could be affective.

If you say you are not very good at something, it may be that you are not concentrating hard enough. What I'm saying sounds easy, it may even sound lazy, but it's not. It requires a great deal of discipline not to be distracted by other things. You have to throw away everything that's irrelevant.


CL: Now I would like to look back what happened this year. What do you think were the good news this year in regard to open source?

Eric

ESR: Well I can think of a couple of pieces of news that I think are of importance. The first one that occurred to me was when it was revealed that Andy Hertzfeld and the rest of the original Macintosh interface designers were going to work explicitly on producing an end user interface, GUI, for Linux. That was spectacular news.

Another piece of spectacular news was the announcement that StarOffice was going to be open source as OpenOffice. It's important because it's going to take a productivity suite with features comparable to Microsoft Office to make Linux truly competitive on the desktop, and Open Office is a very large step towards that.

And the third piece of really terrific news, which has I think received much less notice than it should have. And that is the fact that at IBM, it is now the policy of that company, announced by its president, that every new piece of software written at IBM must be issued under an open source license, unless the development managers can make a specific business case why that should not be. So, that's a sign of the fact that IBM has put itself solidly in the open source camp.

CL: I suspect you may be having a little complicated feeling about IBM committing to open source world, but how do you actually feel?

Eric

ESR: Well, as a board member of VA, they are heavy competition, that means we must work harder and be smarter. As an advocate for the open source community, anyone who wants to get involved, we will welcome.

CL: Would you describe those two feelings in percentage? Fifty-fifty?

ESR: I don't generally have them at the same time, because generally I'm not worrying about VA at the same time as I'm thinking about the whole community. I think that if the whole community does well, then VA will prosper.

CL: Although they differ a little from IBM, Sun is also beginning to have bigger influence on open source world. This year, they bought Cobalt, released OpenOffice, dedicated 50 engineers to GNOME development, and they released their Solaris code under SCSL (it is not an open source license). Now, what do you think Sun's strategy is like?

ESR: That's clearly an anti-Microsoft play. They figured that for a relatively small amount of money open sourcing StarOffice gives them a genuine chance to break a Microsoft monopoly.

CL: Do you think they are really committing to open source?

ESR: I think Sun is a large company with a lot of different people inside, with different views. I've had conversations with many people at Sun, and... let's just say that Sun is still developing its ideas :-)

CL: You listed Eazel as one of the biggest news of this year. Would you give us your comment on the KDE League and the GNOME Foundation?

Eric

ESR: No!

CL: Why?

ESR: Because I am officially neutral in the KDE vs GNOME war and I'm not gonna say anything about those things! That's a political snake pit in which I don't want to wander, thank you, and you can quote me on that.

CL: We saw the big advancement of both GNOME and KDE this year as software and as organizations. But in practice, Linux has not yet been the most popular desktop. Do you see any obstacle to it or is it like it just takes a little more time?

Eric

ESR: I think Linux will become dominant before it is really ready technically for the end users. And the reason I believe that is because I now think that Microsoft monopoly is going to collapse for other reasons in the near future.

One of the reasons is that prices for hardware is steadily dropping. This is a problem for Microsoft because their business model depends on charging a fixed price for pre-installed Windows on a machine. As hardware prices drop, that fixed price represents a larger portion of the margins of the desktop OEMs. The desktop OEMs are going to reach the point when prices drop to a certain level where they simply can't make any money paying the Microsoft tax.

And it is at that point, that the Microsoft monopoly will collapse, because they will then unbundle Windows from their machines, and offer something that's inexpensive (like, say, Linux), in an effort to get some of their margin. I believe that will happen probably within five to six month from now, and that's probably before Linux will become polished and be ready for the end user.


CL: How do you see efforts of open hardware, such as the Open Source Toys Project or MorphyOne as new movements?

Eric

ESR: Open hardware is an interesting idea. I'm a little skeptical about whether it'll work, because I think, the economics of hardware is different from the economics of software. In particular, "release early and often" strategy is easy to do with software, because it basically costs nothing but a little human time to do a software release. You just put together the components, you package them and you ship them. Because it's such an inexpensive process, the rapid iteration strategy, "release often, release early" strategy is cheap and it makes a lot of sense.

This is not true with hardware. Because when you release hardware, you have to push around the actual matter, you have to melt silicon, you have to print circuit boards, you have to put things into boxes and ship them. That's the expensive proposition because it actually involves moving stuff around, instead of bits.

Ruu-chan
(Thanks to LWN's featuring, the Open Source Toys Project is forming an international community now. They have a mailing list in English. Please contact Tomoko Yoshida for subscription. )
And so I think the friction costs for hardware are such that a frequent release strategy of rapid iteration, "try everything and see if that works" strategy, is much harder to imagine that working well for hardware where it isn't for software. So I'm not sure if it will work. However, I hope it will, and I'll watch it with interest.

(We were talking about Tomoko Yoshida of the Open Source Toys Project, who was not there, but is a good Japanese friend of Eric's. One of her students was there with us and she told us that Tomoko was a very good teacher.)

ESR: I'm not a bit surprised.

CL: Have you ever thought of becoming a teacher?

Eric

ESR: What do you think I am? :-) I have spent last three years teaching the world! :-) Not all teaching goes in the classroom :-)

CL: Would you say that being a teacher and being a leader are the same thing?

ESR: I would say that they are not the same thing, but they can converge, they can come together, if they are done in a certain way. Some people lead by teaching, some people lead by doing other things. Some people teach by leading, some people teach by doing other things.

CL: I kind of see a tendency that smart people major in both mathematics and philosophy like you did. Would you explain why?

ESR: Because they're challenging. If you're good at thinking, you will tend to choose the fields where being able to think well is necessary. Mathematics and philosophy are like that.

CL: I think Minsky said that philosophy before computer science was just assumed, not proved.

ESR: I would call that a pardonable exaggeration. An exaggeration but a reasonable one. Minsky was certainly correct in that there are some philosophical questions which are greatly clarified by attempting to build computer models of cognition, which was what he was interested in doing.

But there are other kind of questions which are very important; for example, the particular branch of philosophy I am very interested in is called epistemology, which is a study of evidence and theory formation, the verification of knowledge. Thinking about computers can force you to discard some preconceptions, but it's not necessary to think about computers to do good epistemology... I could talk about this for hours...

CL: You sometimes describe yourself as an anthropologist.

ESR: Yes. An anthropologist is a person who studies the behavior of human beings and societies, who studies how human beings form culture, how those cultures behave, how do they change overtime, how they adapted to different conditions. Most of the important things I've done have come out of considering the culture of computer hackers anthropologically. Focusing less on their technological artifacts and their programs, and more of the social machinery, less on the technological machinery that they used, and more on the social machinery that they created in order to support their behavior, and support their craft.

CL: When I first read your "Homesteading the Noosphere", I was impressed by your tone of analysis, because you yourself were a hacker and your analysis was very objective.

Eric

ESR: In anthropology, this is called "observer-participant anthropology". There's a tradition of anthropologists who are actually part of the culture they described. It's considered difficult thing to do, but it is considered possible. There was a famous case early in this century, in which anthropologist in Africa who was studying African Shamanism, African religious forms. One of his informants was a witch doctor named Black Hornet. Now Black Hornet was an intelligent man, he wasn't scientifically educated, but at least he was an intelligent man. And after a while he started to catch on to what the anthropologist was trying to do, he started to get the basic idea of describing the culture scientifically in such way that other people can understand them. And Black Hornet, the witch doctor became sufficiently fascinated by this. And sufficiently well versed in anthropological technique that he actually collaborated with this anthropologist on several other papers, studying different tribe's systems of doctrine. So that Black Hornet was operating as an observer participant anthropologist within the culture he was part of. So this sometimes happens even with people from relatively primitive civilizations.

He learned that viewpoint from observing the another anthropologist at work. Well, I did the same thing. Not directly, but I read a lot of books about anthropology and about anthropological field work. After a while, I began to see the patterns and began to understand the method they were pursuing.

CL: If it weren't there, I think it took much longer for general people including myself to understand the hacker culture.

ESR: I'm sure that's true. Because it's hard to notice features of the culture that you're not used to. It took a long time for me! I had to be part of this culture for.. let's see I wrote the paper in 1997. I had to be part of this culture for 20 years before I could described it properly.

CL: How did you learn the hacker culture from scratch?

Eric

ESR: By observing the community, people who were already in it. One of the ways I learned was by reading very, very early version of the jargon file, which many years later I became the maintainer of. So, yes, it's very difficult... a fish has trouble analyzing the water it swims in. It's very difficult to notice the features of the culture that surrounds you, which you take for granted. It took me 20 years to get me to the point where I could do that :-)

CL: Did you have to learn a lot before you were able to decide you could manage a project?

ESR: I've watched people do that for a long time. I sort of picked that up unconsciously. I imitated the behavior of people who had done it successfully. Most of the really subtle skills have to be learned by imitation. True of, I think, programming, true of writing, true of martial arts... It's true of most really high level skills. To learn them you need to mimic the behavior of somebody who's already good at something you want to learn.

And not just the obvious parts, but you have to mimic their whole personality, because you don't know what pieces of their personality, what parts of what they are, are important to the skill. So, in particular, if you want to become a hacker, it's valuable not only to mimic the way other hackers program, but also to mimic the language, and the style of humor, and the attitudes towards the world that they have. Because some of that is important and you'll never know what parts that are going to be important in advance. So you need to sort of imitate the whole kit and caboodle, the whole thing, and after a while you'll discover which parts of them are really relevant.

Similarly, if you are training in martial arts, what you're trying to learn is how to throw, punch, kick, grapple - but you really have to imitate a lot more than that. You have to imitate the courtesy, and their wearing of the uniforms, learning the traditional language, because in advance, you don't know what parts are relevant. You have to imitate the whole thing, in order to get their mind set in your head :-)

CL: Would you give an advice for people who might want to start an open source project?

Eric

ESR: Go hang out with your local Linux users' group. Meet the people, soak up the culture, learn what they're doing, get involved in the projects that are already established, learn the ropes there before you launch a project yourself. In software development, most people need to learn how to follow before they can lead. They need to learn how to contribute, and work with the development group, before they can run one themselves. So, go to your local Linux users' group, listen to what people are doing, cooperate with them, help on the projects that are already going on, then after a while, you'll get an idea for project of your own, and you'll know necessary things, techniques, and you'll have the right kind of connections that will work.

CL: Do you see any changes in the hacker community, because the business world joined open source movement?

ESR: No. I don't, actually. I think that most of the people who could be distracted by money were gone long ago, because the demand for programmers has been so strong for so long, that anybody who was not attracted to what we do as a kind of art, has already gone their way to do other things. It's not really possible to change our behavior with money.

CL: Do you see any difference between hackers at your age and those who are a little younger, like Linus or Miguel?

ESR: Well... I have more experience than they do, which makes a difference sometimes. It gives me different perspective. I don't have a lot of differences with them because of my age, we are all hackers together, basically. It doesn't seem relevant very often, it might be relevant if I might go out drinking with those guys, but since I don't drink, it's not an issue.

CL: They sometimes mention that their goal is world domination :-) What would be yours?

ESR: Software that doesn't suck. I think the path to that goal implies open source software becoming the normal thing, dominating the world. But my goal is for software to be of high quality, so that software users don't no longer have to normally expect crashes, hangs, and blue screens of death, and all of the other bad things that comes from shoddily-designed software.

CL: Do you think you have accomplished your goal?

ESR: No, not yet. Personally, I already live in a world of software that doesn't suck, because I've chosen to use Linux for all the things that I do. Most people don't yet live in that world. When most people do, then, perhaps I'd consider my job to be finished.

CL: Do you think Unix sucks, too, like Miguel?

Eric

ESR: I read the actual speech. He was exaggerating to make a point. That's what I think of that comment. He really doesn't think Unix sucks, he just thinks that people need to be kicked in the butt a little bit, about some things that we take too much for granted about it.

CL: But do you agree that Unix has little codebase that can be reused, and this is a problem?

ESR: I think that the fact we have a codebase that we can get a lot of reuse out of is a good thing, not a bad thing. Miguel is trying something very gutsy. He's trying to move a substantial portion of the Unix world from using the traditional sort of interface that depends on text streams and piping and sockets to using a different kind of interface like CORBA and interprocess message passing. The jury is still out whether this is a good idea. Yes, GNOME is doing interesting things, but there's a problem in that the GNOME style of linking programs together is harder to understand, it's less transparent, it's more complex. So, I think the jury is still out on whether that style of interfacing is going to turn out to be better than the traditional Unix style or not.

CL: I was expecting you to say that you absolutely agree with Miguel, because you told Linus about "curse of the gifted" in regard to sharing device driver code. Miguel says "Gnome will actually lower the barrier of entry for hacking on an application project". Would it be that you would prefer KISS ("Keep It Simple, Stupid") to that?

I think it's ironic that the framework of code reuse didn't evolve much (maybe its policy has not been decided deliberately) in free software because we can directly reuse parts of codebase (i.e. cut and paste). But now applications are becoming bigger and bigger...

ESR: I usually prefer KISS, because experience tells me that complex interfaces break more than simple ones. As I said, the jury is still out on this, and will be for a while. Miguel may be right.


CL: Now we would like to talk about bad news. What do you think were the worst news of this year in regard to open source?

Eric

ESR: I have a couple of candidates for that, I can't pin down to any one piece, but, Uniform Computer Information Transactions Act, UCITA, you see it getting passed by a couple of states in the US. It is very bad news. The fact that Amazon actually got a patent on one click shopping, that was very bad news. And the fact that Microsoft stock hasn't crashed, that's bad news :-)

CL: Did you say that Microsoft's stock didn't crash?

ESR: Not hardly enough. It's actually lost a lot of its value of it last year, it hasn't crashed to the point where investors lose confidence and know its business model is broken. That's a real crash.

CL: I see Linux stocks dropping more...

ESR: Well, yeah, the reason for that is pretty obvious. That is, that Linux stocks got caught in the .com crash. To some extent, we were lumped together in perception with a lot of companies that have basically nothing but a concept and some fancy words. When the air finally got out of the .com bubble, a lot of Linux stocks went down with it. It had a very direct effect on VA. The reason we didn't make our numbers last quarter is because some of the .com's that crashed were our customers! We've lost 15 million dollars in handshake contracts within the space of a week!

CL: You can buy VA cheaper now.

ESR: It's a buying opportunity, yes :-) And I can say that, because I'm not in the United States. If I were in the United States right now, the SEC wouldn't let me say that :-)

CL: If some company which is not necessarily interested in supporting open source projects bought VA (like Ajuba Solutions' case), what would happen to, for example, SourceForge?

ESR: There are good reasons for anybody running SourceForge whether it's VA or somebody else. There are good reasons for anyone to keep it alive. One is that SourceForge creates opportunities for consulting contracts to create, essentially, clones of SourceForge for use in development by large software organizations. VA has done quite a bit of that kind of business, with firms that wanted us, basically, to help them create their baby SourceForges internally. And another good reason is that SourceForge enables you to collect a great deal of data about the composition of interests of the open source community, that could be extremely valuable. And it's not like keeping SourceForge going is a very expensive proposition. A few servers and a half a dozen staff, not a big deal.

I think they'll keep SourceForge alive, because they are interested in making money. SourceForge doesn't need money.

CL: Why did VA buy MySQL? Did you have to GPL it?

ESR: We wanted to make sure that there was at least one high quality database that was known to the open source community, that was available under a true open source license.

[Editor's note: VA has not actually purchased MySQL; instead, it is a minority investor in MySQL AB, the system's owner].

CL: What about PostgreSQL?

Eric

ESR: Well, apparently they weren't for sale :-) I don't know, I wasn't involved in the decision to purchase MySQL.

CL: May I understand that you couldn't accept MySQL's old license, or their business model then?

ESR: That's correct, yes. We don't certify licenses with commercial use restrictions (this includes the old Qt, and the Sun Community License, and it could also includes the MySQL license) because it's very hard to define what "commercial use" is.

It's too bad too, because at first sight it sounds like really a good idea to have licenses that say if your software is distributed in a non-profit way, use this, but if it's distributed with profit then you must pay a fee, or accept different licenses or something like that. The problem is, it makes life too complicated. It discourages, it makes people afraid of some kind of behavior which we would like to encourage.

CL: But then, doesn't it mean that the Open Source Definition is kind of library vendors friendly? If it really does, wouldn't it be unfair?

Eric

ESR: I agree it would be unfair. But OSD is not biased. We don't think it is. We don't give people who license libraries any privileges or options that application vendors don't also have. Why do you think so?

CL: Library vendors can put the same version of the same product under both GPL and commercial(proprietary) license. (Commercial developers would buy the commercial licensed one, because they don't want their products to be under GPL.) At the same time, they can also say that their products are OSI certified so they can do PR saying their software is more reliable or more stable or anything that is good in open source development, because its development model is open source.

On the other hand, application vendors cannot do the above, because if they put the same version of the same product under both GPL and commercial license, everyone would choose the one under GPL. (Because, for example, I can use GPL'ed word processor even at work (to earn money, to make proprietary documents).)

ESR: Not necessarily true. Suppose, for example, that they (application vendors) offer support only for the non-GPL version?

CL: Yes, I see that if they do that, they'll be able to get customers who choose non-GPL version.

However, library vendors don't even have to try that kind of disadvantageous business. The reasons I'm saying "disadvantageous" are: a) not all the customers need support for them, b) to do support business, they need different additional cost from software development, and c) others can do the same support business as theirs without paying the development cost themselves.

Eric

ESR: This differs not at all from the library situation.

CL: Library vendors can sell commercial license, and so they do not have to get themselves do support business. This is unfair.

ESR: I don't see it as "unfair". These two groups are in different businesses, they take different risks. That's normal.

CL: I think it'll be more easily accepted by commercial vendors if it is possible for them to sell not only support but also commercial licenses.

Certainly MySQL's old license may have had small defects, but what I think is important here is that charging money when commercially used (aside from legal definition problem) is basically a reasonable idea (business model) that the open source community actually can accept.

ESR: Except that the open-source community does *not* accept this idea. OSD's prohibition against commercial-use restrictions has never been controversial.

Eric

CL: In fact, library vendors can have a way (loophole) to both be OSI certified and be earning money by licensing their software the same time. (i.e. OSD does not prohibit dual licensing under which non-open source license can be also chosen.) So practically speaking (or, as a result), the only difference between the two (specifically, current Qt license (business model) and the old MySQL license (business model)) seems to be either they indirectly prohibit commercial use, or directly (explicitly) prohibit it.

Since both OSI and FSF are interested in (or focusing on or aiming at) making sure that code is freely available, and not in helping particular vendors, I'm sure it is not intended to be unfair, it's just a result. It's easy to say just no to it, but I think it'll be better for the community to somehow hack the way (invent the way both the community and application vendors can be satisfied with) to be able to say yes to it.

The way library vendors and the open source community get along with each other is already hacked (invented). I think there could be a way for application vendors too but I cannot think of any specific idea. Do you have any good idea?

ESR: Yes. Do what Sendmail, Inc. does. Use your application as a library-like basis for proprietary add-ons.

CL: Have you heard of this kind of discussion before?

ESR: Yes, but generally, most other people don't care about these details as much as you do. I'm not saying that's bad, I'm just saying that most people don't.

I have these kind of discussion with Richard Stallman, occasionally.

CL: You and Richard Stallman talk about license!?

ESR: Of course we do! It'll be crazy if we didn't!

CL: Would the talk end up in fighting?

Eric

ESR: You're not aware that Richard and I have been friends, for twenty years? There are little tensions sometimes, but we do talk. There wasn't any tension before I got famous. That made things a little complicated.

CL: Do you take licenses one by one when you talk about licenses at OSI?

ESR: Yes, we sometimes refer back to licenses that we've discussed previously, in order to clarify our thinking, but we're taking these one by one.

CL: Do you find thinking about licenses interesting?

ESR: It's not especially interesting to me, it's just necessary. It's not something I would choose to do unless I considered it very important. But since I do think it's very important, I worry about it.

Eric

CL: I think thinking about licenses is like thinking about business models, so I find it very interesting.

ESR: Oh, that's interesting. But that's different from worrying about licenses. Thinking about the policy of what license issue you would accept is an interesting process of trying to clarify the values of the open source community and what's considered as essential. That's at a higher level from parsing the language doing the analysis on specific licenses, which is often rather boring and tedious. So those are kind of different questions. Do I enjoy thinking about the ethical issues that are underneath? Yes I enjoy thinking about those. Do I enjoy parsing and analyzing the logic chopping of specific licenses? No, not really. There are other people who are much better at it than I am.


CL: You listed software patents as bad news. Would you give us more specific comment on this issue?

ESR: The one-click shopping thing? (sigh :-) It makes me wonder what drugs the patent examiners are on.

I think I can imagine a software patent system that was fair and equitable, but it doesn't resemble the one we have at all. For one thing, I think any reasonable software patent system would have to accept the entire body of open source as prior art, so if the technique was created as open source then it couldn't be patented.

Secondly I think that independent re-discovery of the technique should be an affirmative defense against patent infringement, that is, you can't patent a technique, keep it secret, and if then somebody else uses the same technique, suddenly surfaces with the patent and say "oh, here's our secret technique which we have patented, and this means you can't use it." I think independent re-discovery ought to be a defense against patent infringement. Right now, it's not.

Eric

There are a couple of other technical points about the way patents work which I think would have to change, but clearly (beyond these problems which relates to the patents system that might work in the ideal world) there's also an issue that the software patent examiners we have are incompetent. From the patents they have been giving out, they're clearly not equipped to distinguish original inventions from things that are obviously the state of the art. And further more, the systems of incentives under which they operate is skewed in favor of patent applicants, because the examiners get paid when they issued patents and they don't get paid when they deny them.

So, what you see is a mixture. You see, there are some software patents which I think are certainly legitimate, because they represent interesting and novel work that does not occur in any prior art. For example, a patent which I would consider valid is the RSA patent which just expired. That was an genuinely brilliant original work that deserved a patent. Excellent example of a bad patent? Examples are legion, but the Amazon's one click shopping was bad, the Lempel-Ziv patent that blocks the use of GIFs, that's a bad patent. British Telecom holds a patent on hyper-linking, which they're actually trying to enforce, that's a bad patent.

So clearly, at minimum the system has to be reformed, so that the novelty standards for software patents are much much stricter. And then, there are also these other changes, which I discussed before.

CL: What would you say about the idea that OSI and FSF do not stand against them but do help and encourage free software developers to obtain patents, and make those patents available only for use in free software?

GPL is great in that unlike the idea of Public Domain, it emphasizes on copyrights and makes sharing of *code* among free software possible by means of licensing to make it sure that free software keeps being free software.

But GPL treats the *ideas* the same way as Public Domain does. In other words, even if a new good idea was proposed and implemented in free software, proprietary software can utilize it freely, and the opposite is not necessarily true. (ie. They prevent open source community from using their ideas, by obtaining patents.) This is unfair.

Eric

ESR: Well, it's a problem, but I'd say either we believe in the effectiveness of open source development and that will spread because of economic advantages, or we don't. If we believe in the effectiveness of open source development, then it will eventually become normal in spite of patent problems. If we don't believe their development effectiveness is effective enough to mostly displace closed source software, then I don't think the patents are going to make a critical difference.

Which is not to say that I don't think the patents are not problem, software patents are a very annoying, irritating problem. But I don't spend a lot of time worrying about them, because I think we've got larger issues to be concerned with, like, demonstrating repeatedly often enough to convince conventional software development model managers that they are viable business models that don't depend on code being secret. If we can't make that demonstration, what happens with patents doesn't matter. If we can make that demonstration, then patents will eventually cease to be a problem.

Eric

CL: Within this century, we have come this far as to be able to set up a modern desktop environment without any technique that is software patented (except those that are re-invention of prior arts). But I wonder if this holds true for the next century too. Much more companies are eagerly trying to obtain software patents now, than they used to be. I have no doubt that many of the currently developed infrastructures will be applied for patents, like those of high quality motion picture or natural language processing, etc.

If I were asked by my family which one to use, Windows 2010 with all of those technologies of the 21st century, or Debian 8.2 (codename:"pomato") without them, I won't be able to recommend Linux to them.

So I think open source community should start obtaining patents of technologies that we ourselves invent now, so that we might be able to cross-license with others including proprietary companies. I think it's going to make a big difference.

ESR: You may be right. But I can't solve every problem in the open-source world. I have to concentrate on one battle and win it. Battling software patents will have to be someone else's problem, I'm afraid; I must concentrate on the parts of education and advocacy that I do well.

Eric
(Eric answered our questions in spite of the strain of traveling during the interview, and he also answered our questions by email. Thank you very much, Eric! We hope to see you in Japan, again.)


-- Maya Tamiya and Masa

Eklektix, Inc. Linux powered! Copyright 2002 Eklektix, Inc. all rights reserved.
Linux ® is a registered trademark of Linus Torvalds