Live Free or Die

Since heading up the Open Source Initiative, Eric Raymond has become one of the most prominent advocates of open source software.


If Linux has a public relations department, it’s Eric Raymond. Long-time hacker (he wrote fetchmail) and self-confessed cultural ethnographer, Raymond has had much to do with the intense media buzz that has sprung up around Linux and open source software in the last few years. As president of the Open Source Initiative, his essays and public pronouncements have done much to express the open source value proposition, and his infamous leaking of internal Microsoft documents about open source (the Halloween documents) made it clear that not just the hackers were listening to what he had to say. Raymond recently showed Linux Magazine Publisher Adam Goodman and Executive Editor Robert McMillan around the offices of VA Linux Systems, where he has a seat on the board of directors.

LINUX MAGAZINE: Maybe we should start at the beginning. Where did Eric Raymond spring from?

ERIC S. RAYMOND: I’m actually a second-generation hacker. My father programmed from the earliest days of the great mechanical dinosaurs in the 1950s into the 1960s — Univac machines. He learned computers in the Air Force. He was involved with high-altitude meteorology and he was always very proud of the fact that he was the first enlisted man to take the Air Force’s computer-training course.

LM: So you grew up with computers in your blood?

ESR: That’s right. When I was 11 years old and living in Rome, I got to play little wireframe video games with an $8 million mainframe computer. Nobody even knew what video games were in 1968.

LM: What was it like seeing a computer for the first time?

ESR: I was fascinated. Like my father, I’m wired to be a techie. Complicated machinery fascinates me. For a while I thought I wanted to be a pure mathematician, but then I burned out on that and taught myself computers; I’ve been programming ever since.

LM: Where are you from, originally?

ESR: I was born in Boston, but I’ve lived in or around southeastern Pennsylvania longer than I’ve lived anywhere else, and it was at Penn [the University of Pennsylvania] that I got my first exposure to the Internet back in ’76. Back when it was the ARPAnet. When I got my first exposure to the hacker culture I was fascinated by it. That’s where I saw the Jargon File for the first time. I think I first saw it in 1977, and I was absolutely fascinated by it, because it showed me that there wasn’t just a technology here but a whole culture.

In a way, everything I’ve done from the Jargon File through the Cathedral and the Bazaar on down has been working out that fascination I felt the first time that I saw the Jargon File.

LM: And how did you get into Unix?

ESR: I guess it was 1974, I read an article about Unix in the proceedings of the ACM [Association for Computing Machinery] or something like that. My father received this as an ACM member. I didn’t understand what they were talking about, but I sensed that there was something special here. There was a kind of design elegance that I hadn’t seen anywhere else, so I was actually a Unix fan before I was a programmer.

Eric 3

Eric 2

Eric 1

LM: A lot of people suggest that there are similarities between the Linux culture today and the early Unix culture.

ESR: Oh yeah. There’s a similar feeling of tribal spirit, a similar ethos of sharing code, and a similar sense that: “Hey, we are doing the best technology.” The difference is the Unix culture was slower. We didn’t share code as effectively. We didn’t really have the Internet yet.

For years we only had poky UUCP [Unix to Unix Copy Program] connections. It really wasn’t until the Unix culture mated with the Internet culture that the synergy really started to happen.

LM: Was it really a matter of TCP/IP or do you think that it was a product of the time as well? The long hair, beards, that kind of laid-back attitude? The ’90s are much different.

ESR: Let me tell you about a change I’ve noticed in the hacker culture. Some people find it hard to believe, but J. Random Hacker is actually rather healthier and better adjusted today than he was in 1978.

LM: Why is that?

ESR: I don’t understand completely. It’s partly because of the culture. I think we used to attract certain really desperate, marginal people and now, because the culture’s moved towards us, we’re attracting people who aren’t quite as far out on the social-adjustment bell curve.

LM: I think there’s a cachet to being a geek these days.

ESR: There wasn’t back when I started. You kids. You have it so easy!

LM:Rolling Stone was around in the 1970s. Can you imagine a Unix hacker being inside Rolling Stone?

ESR: Noooo. It would never have happened. I get a really strong sense that it’s a lot easier to be a bright kid in high school today than it was when I was in high school in the early ’70s. This is a culture that actually respects intelligence. The kind of gut hostility to it that I used to experience has abated somewhat.

LM: You could fry Bill Gates all you want but I guess he’s made geek chic?

ESR: I hate to think he had anything to do with it. That is a distressing thought.

LM: I think that money validates whatever you’re doing. People respond to money.

ESR: What an awful thought. I feel kind of disgusted.

LM: Okay, enough about Bill Gates, then. What about Richard Stallman? You’ve known him since the early days. How did you two meet?

ESR: I met him at a science-fiction convention in late 1976. At the time he was working on Emacs. Emacs hadn’t actually seen the light of day when I first met him, and I remember one of the first conversations I had with him was about this cool software that he was working on. If anybody had told me that a good portion of my life would actually turn out to be bound up with that software, I would have said they were crazy.

LM: Did you get along when you first met?

ESR: Yeah. We were pretty friendly.

LM: Did you bond on the free software thing?

ESR: No. The free software thing didn’t happen until after we’d been friends for a while. In fact, there was a period of about a year-and-a-half there when I was hearing legends about this guy named RMS and I failed to connect him up with the Richard Stallman I knew.

I remember there was a moment some time in probably 1981 or early 1982 that I finally realized: “Wait a second. I know this legendary geek. He’s Richard! The guy from the science-fiction convention!”

LM: How has your relationship changed over the years?

ESR: Surprisingly little, I think, because I always argued with him a lot. The difference is that ten years ago our arguments were private and nobody cared. Now they’re public.

LM: What are the differences between Open Source software and Free Software?

ESR: I think intellectual property is not a bad thing. Actually, I think intellectual property is a fine thing. I think the decision to give up your ownership rights in a program or a book or any other kind of intellectual property shouldn’t be based on somebody making a moral claim that owning that kind of thing is wrong.

It should be based on questions like: What’s economically effective? What leads to good engineering outcomes? How do we get the best possible results?

LM: So where do you think your and Richard Stallman’s philosophies merge? What do you guys have in common?

ESR: I think we both care passionately about the art of software and about the culture that we have in common.

LM: It’s interesting that you don’t mention freedom because I think of both of you as people who care about freedom.

ESR: Yeah. We have pretty different definitions of it, though. I’m an individualist; he is less so. I believe freedom ultimately comes down to the right to be left alone. The right to enjoy your own life and enjoy your own property and deal peacefully with others and not be messed with by people with agendas.

LM: Would you say that a lot of the terms of the GNU GPL in some ways restrict freedom, then?

ESR: That’s not an argument I’m interested in having. Some people will choose to use the GPL because it fits what they want to do. I use it sometimes. Some people will choose to use other licenses because the particular kind of social engineering or community they want to set up is different. The right tool for the right job. I’m not religiously pro-GPL; I’m not religiously against it.

LM: But Stallman would argue that you need to have certain restrictions of freedom to prevent greater restrictions of freedom. Like you have to say: “All software must be freely distributable in order to ensure that nobody can take that away from you.”

ESR: I’m not interested in having that argument with him. I think we have an entirely sufficient justification for open source and the GPL: solid pragmatic grounds.

LM: What is going on with the Open Source Initiative (OSI)?

ESR: We’re doing our job.

LM: Do you have meetings?

ESR: We have e-mail meetings. We have day-long e-mail meetings and mostly what we do is, various members of the board give consultation and advice to corporations who would like to play the open source game so that they can do it honestly and cleanly, and we process requests to certify licenses as being OSD- [Open Source Definition] conformant.

What we’re trying to do is educate people outside our community about the social contract that knits together the open source world. We teach them how to play nice.

LM: When you launched the OSI, it seemed that it was going to be the branding body for open source software. You were going to be a kind of watchdog group that would issue a certification mark and…

ESR: And that’s what we’re doing right now.

LM: What programs are in place to make this happen?

ESR: IBM and Apple have both come to us for license certification. I expect we’re going to certify Python shortly. The list of certified products and vendors on our products page has steadily grown over time, so we’re doing pretty much what I originally envisioned we would do.

LM: How do you typically work with a company like IBM or Apple? Do you just set up shop and wait for them to call you?

ESR: That’s right. They send OSI an e-mail saying: “Hey we’ve got this license we’d like to get certified so we’re properly conformant. What’s your process for that?” We then look that license over. We have a license evaluation specialist — a person named Seth Schoen. He’s very well known for being absolutely brilliant as far as the legal language goes.

Generally, after a few iterations we come up with something everyone is happy with, and then we certify that and the corporation releases it.

LM: How did Netscape become an open source company?

ESR: Well, it all starts with Tim O’Reilly. We happened to sit down next to each other at Linux Congress more than two years ago. I had wanted to meet him for a number of years because I had heard of O’Reilly & Associates for a long time.

You can’t study that company for very long without noticing that it seems to be the creation of a single mind. It has a kind of consistency about it, and it’s clear that that mind is a pretty interesting one. So I really wanted to meet the mind behind that organization.

When we began talking, we discovered that we had a common interest in classic science fiction. We spent about three hours talking about everything under the sun. By the time that conversation was over, we were friends. So that led to him inviting me to speak at the first Perl Conference in 1997.

Apparently some people from Netscape were at that conference, and they heard my talk and they went: “Wow!” Then they took that back to their bosses at Netscape. So, basically, Barksdale picked up on the paper because of what the people in the Perl Conference told him about it.

LM: So who brought it to Barksdale?

ESR: I don’t know. On the 23rd of January 1998, I get a piece of e-mail from somebody who says: “Eric, I think you better go look at this press release.” So I go look at it and it’s their announcement that they’re going to open source the browser.

Of course, open source doesn’t exist yet — neither does the term Mozilla — but they’re announcing that they’re going to release the source code and as I’m reading the press release, it looks kind of familiar. The language seems suspiciously like something I have encountered before.

So I called them up and I left a message asking: “Will somebody please tell me if I had something to do with this?” They called me back and they said: “Yes you did.” That led fairly directly to my getting asked out there to consult with them on how to set up the Mozilla organization. That was in early March, 1998.

It was while I was out there in March that the nucleus of what later became the Open Source Initiative got together and we coined the term “Open Source.”

LM: Who came up with the term?

ESR: It was invented by a woman named Christine Peterson — the executive director of the Foresight Institute — a think tank in Palo Alto. They’re concerned with nano-technology. Their take on the situation is: “Hey, if you’ve got nano-bots running around in your body, the software that’s running them had better be reliable.” So they’re really into software reliability.

LM: It’s been over a year since the Halloween documents were released, and there were a lot of things last fall that really accelerated the mindshare of Linux and open source software. The Halloween Documents were key. What are you thoughts on the one-year anniversary of their release?

ESR: We’ve been able to use them to force Microsoft’s FUD mongers to keep their heads down because anytime they pop up saying negative things about Linux, we’re going to wrap all the stuff they said in the Halloween Documents around their heads like a tire iron.

LM: Has Microsoft actually learned anything from the hacker culture?

ESR: Of course they’ve learned a few things.

LM: What have they learned?

ESR: I don’t know. I’m sure they’ve learned something because they’re not stupid, but I couldn’t enumerate a list. I don’t pretend to understand Microsoft that well.

LM: Did you always find Microsoft to be an ominous entity?

ESR: I view Microsoft much the same way in which I view government. Colossally stupid, occasionally quite evil, and to be abolished as soon as possible.

LM: You really feel that negatively towards them? That they’re stupid? I mean, a second ago you thought they were pretty smart.

ESR: Individually they’re smart but their corporate culture is stupid.

LM: Do you single out Microsoft as a proprietary software vendor, or do you also think that Sun’s proprietary software sucks or that all proprietary software sucks?

ESR: I’ll say this: All software that hasn’t been thoroughly peer reviewed has to suck pretty hard. You can draw your own conclusions from that.

LM: Are you at all concerned about the success the open source movement has been having? When Red Hat went public a lot of people were very excited, but at the end of that day, I remember at LinuxWorld a lot of people were kind of looking at each other, realizing that things were not going to be the same anymore.

ESR: Hey, this is success. Cope! Success may be hard to deal with but it sure beats failure. I would rather have the problems of success than the problems of failure.

LM: What do you think the problems of success will be?

ESR: Too soon to say.

LM: There’s nothing on your radar at this point? The community seems to be capable of coming up with lots of them.

ESR: I could repeat all of the obvious clichés: People will be distracted by money, blah, blah, blah, blah, blah. It’s too soon to say. I’m really reserving judgment on that.

LM: That’s interesting. So dealing with success is really not on your agenda right now? You don’t feel that it’s totally been achieved at this point and, for now, you’re still focused on getting there?

ESR: Right. And even if I thought it had been achieved, there’s not much I can do as an individual to solve those problems. That’s a growth process — an adjustment process — that the whole culture is going to have to go through.

Sooner or later, they’re going to have to face up to the question of what their deepest desire is. I’m not worrying about that because I know what my deepest desire is. I figured that out about 20 years ago.

LM: What is your deepest desire?

ESR: Well, I have more than one. I’ve been quoted as saying that I want to live in a world in which software doesn’t suck. That’s really important to me. It bothers me that a lot of software engineers — guys like me who want to do a good job, who sweat blood to do a good job — have to sit up at 3 o’clock in the morning in bed wondering: Why is it all so rotten? That really bothers me.

LM: A few years ago many of the top Linux hackers were basically working for nothing, but today they’re all employed by soon-to-IPO Linux companies. Do you think that the new, corporate open source cultures that are evolving are going to be able to nurture the next generation of great hackers?

ESR: Too soon to tell.

LM: Do you think that’s possible?

ESR: It wouldn’t surprise me. I think the next generation of hackers may nurture itself. The same way it’s always been. I think that hackers basically create themselves. If they’re dedicated enough, they’ll find a way to hack, regardless of what the economic conditions around them are. There’s even a hacker culture in post-Communist Poland, which is a very poor country. They still haven’t recovered from being bombed in World War II.

LM: So what’s the roadmap for Linux, as far as you’re concerned?

ESR: First take the Internet, then take the business data centers, then take the desktop. We’re most of the way through taking the Internet, we’re getting a good running start on taking the servers, we haven’t done much with the desktop yet.

LM: Do you think that taking the desktop will require Microsoft Office?

ESR: I don’t know yet. It’s too soon to tell. But it could. Or some clone of it. I’m very encouraged by what the AbiSource people are doing.

LM: How do you see them as really making a profit though? There’s really not much service and support for Office still.

ESR: Microsoft might argue with you on that.

LM: Is there any danger of the Linux distribution fragmenting?

ESR: What Linux is actually doing now is absorbing the old line Unixes. Not destroying them — that’s significant. Because it’s not taking marketshare away from them. It’s simply engulfing them. Engulfing their user base and their technical base.

LM: I agree with that but, to give you an example, SGI is now getting into Linux, and it seems like they’re almost trying for the SGI brand of Linux.

ESR: That’s okay. They can do that if they want. Fragmentation happens because the source is with the pluralist. It happened because the incentives were to differentiate the product in order to capture more advantage from your supposed, wonderful IP. In the Linux world, you can’t play that game.

LM: But SGI could still be different from Red Hat?

ESR: That’s okay. We already have that. It’s called Debian.

LM: When you start talking about hardware, though, if you have SGI differentiating from VA Linux Systems or from basically optimizing the distribution to run on a certain hardware configuration that is not available: What then?

ESR: So how is this different from a PowerPC play? How is this different from a StrongArm port?

LM: At the level of application compatibility. You have to do things to make software for a StrongArm port run on Intel.

ESR: It’s called a recompile.

LM: Ideally you wouldn’t need to recompile between Intel boxes. You don’t think that’s a problem?

ESR: I don’t see a problem here.

LM: Perhaps it’s more of a psychological issue.

ESR: Well, if it’s a psychological issue, we aren’t going to make it go away by obsessing about it in public, which is one reason I want to cheerfully dismiss it.

Robert McMillan is executive editor for Linux Magazine. He can be reached at bob@linux-mag.com.

Comments are closed.