GNUisance: Pigdog Journal Interviews Richard Stallman

ES: One of the criticisms I've heard of Free Software is, um, that it's too hard to use. Do you think that eventually there'll be, like, nice GUI applications?

RS: It's clear. You know, given that two different projects are working on such things, it's gotta happen.

ES: Would Mozilla be one?

RS: No. Gnome and KDE.

ES: Gnome? What about GNUStep -- GNU OpenStep?

RS: GNUStep I think is sort of --

LE: It's a framework.

RS: It's a framework, and I guess it, you know, I think it interoperates with Gnome now, and there's lots of people on that, so...

NM: I've been using WindowMaker for a long time now. That's, that's the window manager for GNUStep. And I know that there's a page called GnomeMaker, which is how to get Gnome and WindowMaker together, so... It proves that they are working together.

RS: So, uh, I don't if WindowMaker is build using GNUStep. I think it just works with GNUStep.

[General assenting noises]

But I think that uh, I think that uh, GNUStep works on their stuff to make it work with Guile and interoperate with Gnome. I think. I'm not sure how far they've gotten. Uh, I don't give, uh, I don't give GNUStep as our focus. It's something some people work on because they like NextStep and they want something like that on top of... I'd be happy to have them working on it, but, uh, that's not the GNU project's desktop focus.

Gnome is the project's desktop focus.

RM: There's a current editorial by Jim Dennis urging the Gnome and GNUStep people to join forces in some fashion. I haven't read it yet. [Then maybe you should pipe down, Rick!]

RS: Wholly I agree. I've been talking to the GNUStep people all the time, saying it's important to work well with Gnome. And I, I don't remember exactly what they said, but my vague memory is that they said, "Yeah." So I don't think that there's any need to, to cajole people. I think that they are working in that direction already.

RM: You may find it to be of interest, by the way.

RS: Yeah. [Does that thigh-drumming thing again.] But yes, you're right, it's, well, when the GNU project started out we set out to make a system that was like Unix. And that's what we got. The GNU/Linux system is a Unix-like system. It's as powerful as Unix, and it's as easy to use as Unix. [Laughs]

So we need to make it easier to use. Free Software shouldn't be only for hackers. It should be for everyone. It's not surprising that the first thing we made was for hackers, was being made by hackers. But fortunately there do turn out to be hackers interested in making things that are good to use for non-hackers.

MB: Right. Which is one of those kind of arguments that people use against Free Software. That hackers will never make something that's not for them.

RS: But what's happening is that nowadays the events are proving them wrong.

NM: Now this is, this is partly where I tend to think that proprietary software for Linux [Damn him! He got away with not saying "GNU/Linux"!] can be an interesting means to more Free Software for free operating systems.

RS: No, of course, it's not for Linux of course, it's for GNU/Linux. [BUSTED! Ha ha]

NM: I was thinking of the... OK, yeah, now that we're using glibc, I was thinking more the...

RS: It always was.

NM: You're right.

RS: The C library that was used in the GNU/Linux systems was always a modified version of the GNU C library.

NM: But, um, it's sort of like, you know, people get used to saying, you know, people get used to saying, "Well, you know I've this great free GNU/Linux system, except for this one program that I have to use to fulfill one need or another." They use that, and then they say, "Well, now it's time to write a free replacement for it." And I've seen a lot of that.

I mean, I mean, there's been apps out there for a long time, and people are saying, "Well, you know, we really do need a good, FREE word processor of this quality (or better, hopefully) soon," and that's what the OmniSource (?) people are all about.

Sort of like --

RS: Wai-wai-wai-wait a second! This is not what you're saying! What we're seeing here is that A) there was a proprietary package available and B) somebody's writing a free package.

NM: Right.

RS: I don't the existence of that proprietary package was necessary or -- there's no reason to think it played a specific role in convincing somebody to build a free one. It's not as if -- what you said before is a person is using a proprietary one, and then he or she decides to develop a free one. It's not clear if that is happening here.

NM: OK. You're right, you're right.

[Garbled arguing -- RS and NM are talking over each other. I can't decode it.]

RS: --are former Applix users. Right?

NM: Right. OK. But I can see someone getting fed up with the walls around a proprietary program, that one or two projects --

RS: You CAN, but --

[more garbling]

RS: I think we could have an even stronger response if we all had the idea that proprietary programs were just out of the question. Just off limits. Then the motivation to write a free one would be even stronger.

The number of people who *don't* write a proprietary one -- [correcting himself] who don't write a free one because there is a proprietary one that they can use, if they can morally stomach it, what that does is mainly two things.

One, the motivation of "I want to get this job done" goes away as a motivation to write a free program. And two, the act of using a proprietary program leads you to find excuses to say it's OK. Because otherwise you'd feel uncomfortable with yourself. So what it does is it undermines the resolve of people against proprietary software.

So I think that it does mainly harm in two, in two different ways.

NM: OK... See, the thing is, my job -- part of my job -- where I work, at S.u.S.E., is at least internally replacing proprietary programs that we've been using internally from way long ago with, y'know, Open Source replacements.

RS: Yeah, on the other hand, S.u.S.E., S.u.S.E. is known for including lots of proprietary software in their GNU/Linux system.

MB: Well, that's because A) they're Bad People and B) they're Germans.

[Laughter]

RS: I'm not gonna say that.

MB: That was me. [into microphone] That was me.

RS: I don't want to say a bad thing about Germans. Uh, you know--

MB: That's OK, because I'll do it for you.

RS: It's been about sixty years since there was a reason to do that. And, uh, I have no reason to think it's because they're German. There are lots of idealistic Germans working on Free Software.

But this company has chosen a business strategy that might be an effective business strategy, but it's not based on concern for freedom. And what can you expect?

You know, when you've got, when you let decisions about freedom be made based on somebody's business interests, what do you get? You know, it's rare when you get something good.

I think it takes a certain amount of idealism that most businesspeople don't have. You can make a business that's ethical, but there's constant temptation, especially if it looks like whoever does the unethical thing first and most is gonna have an advantage over the rest.

NM: You also have to look at the fact that S.u.S.E. has done some other things as well, namely in the XFComServer (?). In other words, they wrote proprietary X servers for chipsets that weren't supported, and worked as hard as possible to be allowed to make them free.

RS: And were they allowed to make them free?

NM: The Geometrix (?) server is now. There's one other now that's free.

ES: The elsub (?)?

[Gar gar gar]

LE: The credit for the gar gar gar server is given to Red Hat.

NM: Really?

LE: I was just reading that the other day and I was really surprised.

NM: But I don't believe they ever wrote one!

RM: I believe that's an error.

NM: Really?

LE: I'm pretty sure I saw it last night while I was reading through all the documentation.

RM: Let's just say that there are several X servers -- including in XFree86 3.3 -- that got their start with S.u.S.E. GmBH [He really said "GmBH"!] as closed-source--

RS: --proprietary--

RM: --proprietary products, under NDA's from the chipset manufacturers, who got gradually coaxed by S.u.S.E. and other people to allow first the replica-- putting out the binaries, of course, and then making the source code actually free.

The remaining stuff that has yet to be done, in I think all of these, is convincing them to open up the APIs so that the now-free source code can be effectively maintained. It's really rather hard to debug the source code, however free it may be, if you don't know the APIs.

NM: However, it has made some interesting steps for some people to reverse-engineer and come up with their own lists of APIs. Reversing the code. Which is kind of an interesting backwards way to look at the hardware.

DM: [?] certainly has given the world a lot of hard practice doing that.

NM: Oh boy.<'p>

RS: See, this can be done by, you know, by reverse engineering a proprietary XServer, or any proprietary driver for the hardware. There's another way to find out the specs of the hardware.

I believe it's now legal in the U.S. officially...

MB: Is it? To reverse engineer a --

RS: For that particular purpose, of making something that interoperates.

MB: So, but --

RS: According to what I read that was an explicit provision in the recent mostly-disastrous copyright law.

NM: That's another thing that I always wondered about. I remember reading something on a Web site about this, but I almost always that that what would have been more useful at this current point in time is not necessarily freedom of software -- software free to be modified and distributed -- but simply the legal permission to write a particular kind of software. I mean, people aren't allowed to write certain kinds of encryption software --

RS: Well, listen listen listen. It's true that people are not allowed to write certain kinds of software. But the reason is, not exporting, what you're thinking. The reason is a patent. You can't implement the RSA algorithm.

Patents are a very big threat to the Free Software world. You can't write MP3 encoders or decoders. There are no, there is no Free Software for that, and it may be 15 years before there is. I don't know when those patents were issued -- I think it was 5 years ago or more, so we probably have 15 years or less to wait. That's a *long time*. It's really sad. You know, it may be obsolete by the time we're allowed to use it. And something else will be the rage that we're not allowed to use.

Software patents are really dangerous.

LE: Could you FTP a file to a MP3 compression server in a country where software patents are not recognized, and then download the resulting MP3's?

MB: Are there countries where software patents are not recognized?

NM: Anguilla is the one that Seth David Schoen always quotes to me as the ultimate anarchic paradise. I'm not sure whether or not that's true, but I do know that a lot of interesting work sort of goes through there.

There's this guy, his name is -- his user name is -- n@ai, because his name is Ian, and that's Ian backwards.

"Dandy" Don Marti, Wannabe Hakr d00d

RS: Well, anyway. You know, you might do it. I'm not sure. I don't know what patent law says about that. But you know what else you could do? You could just post an encoder there, and in fact as a practical matter people could snarf it. But you could get sued. DM: I have a decoder on [garbled]. The original developer had to take it down cause-- RS: Well, it should be, it should be posted in Anguilla or someplace else where you can't get sued for that. DM: Well, I think if I make their lawyers spend 32 cents on me, and they spend 32 cents on everyone else who has a random software directory, then it might slow them down a little.

RS: Of course, what slows them down even more is post it in Anguilla. If it's on the Web in Anguilla, and therefore anyone who wants a copy can get it from there, and these lawyers are never gonna know who it is.

More Stallman

 

dunsmuir@pigdog.org