Visit Your Local PBS Station PBS Home PBS Home Programs A-Z TV Schedules Watch Video Donate Shop PBS Search PBS


Main NerdTV Page NerdTV Archives
send feedback rss/podcasts I, Cringely Sponsored by Datapipe Creative Commons License

NerdTV #5: Tim O'Reilly

Bob: Tim O'Reilly, welcome. I keep wanting to speak in an Irish accent. I don't know why.

Tim: I'm not that Irish.

Bob: I, Tim O'Reilly. Welcome to Nerd TV.

Tim: Thanks for having me.

Bob: Okay great. Tim, how did you get in the computer business?

Tim: Well, I still remember the day I saw my first computer. It was the same day that I got my first job in the computer industry. I had a friend who had been a programmer. I met him through; he was taking a class that my wife was teaching on non-verbal communication. And I was a writer. Just pretty fresh out of college. He was looking for work. He got offered a tech-writing job. And I said I'd back him up. And so the first project I did actually I did sight unseen. I helped him write a--it was called the LPA 11K Laboratory Peripheral Accelerator User's Guide. The Digital Equipment Corporation. And I was just basically an editor and kind of made his prose work.

But then it worked pretty well so we decided to apply for a job together. And it was up at the Digital Equipment Corporation's you know Nashua plant. And on the way we stopped by a digital office in Merrimack, New Hampshire where he had previously done some contract programming. And he showed me a computer and we went in and applied for this job together. And we pitched them on this idea, which actually was fairly formative for the company. That what we could do was you paired up a programmer with a writer and you got better results. And for whatever reason they were desperate enough they hired us both when they were just looking for one tech writer.

And that was really our pitch for our first you know four or five years. It was a partnership and it actually worked pretty well because I didn't know anything about computers and so we'd sit there with the engineer and they'd spat off stuff that I couldn't follow. And then we'd go back to our cube. And I'd say so tell me what they, explain this to me. And then I would write it down. Problem was my partner never did learn how to write. And I eventually did learn about computers. And therefore---

Bob: And then you had to kill him.

Tim: Then I had to get rid of him. It was just, it was sort of tough because we took on more and more work and I was. We'd get into these big fights about who was really doing all the work and it wasn't until we hired another employee, our first employee and she was around for one of our fights. And she said, well, she took me aside afterwards and said, "Well I've only been here for a month, but I could tell you that you are doing all the work." (Laughing) And so we then we agreed to split up and I never looked back.

Bob: Now when your wife's teaching a course in non-verbal communication, is she allowed to speak while she is teaching it?

Tim: Absolutely. A lot of what the work that she used to do was really about getting people to be more aware of sometimes what they were feeling. Or being more in touch with what their kind of what their inner was saying at the same time as their outer might have been saying something different.

Bob: Is it freaky to be married to her?

Tim: No, not at all. It's actually we met through this work. Actually in my first public speaking was at Esalen in the late '70's. So, early '70s rather. So I used to leave workshops with her and this guy that I'd study with when I was a kid.

Bob: So you for a long time you were just faking this.

Tim: Yes and no. I mean in the sense that first of all with computers a lot of technical writing is really about seeing patterns. I still remember in those early years I did some of my best work because I didn't really know what I was talking about. And so I was, what I would do is I would just read the specs over and over again until I could see the patterns in them. And I remember once on this one manual and it was basically a lot of times when you are writing computer manuals, you get all kinds of raw material. You get specs. So it's really an editing job and an assembly job as much as anything else. And figuring out what stays and what doesn't. What tells the story, how do you kind of pull it together?

And there was some manual and I said, "Wait, all these three things are actually about the same thing from a different angle, you know? And I kind of totally threw it all out and rewrote it." But I would often find that I could see the structure better even when I didn't actually understand the details. And of course the programmer had written down the details. So I would just have to preserve them.

But I would you know be able to try to figure out how to tell the story in the right order. And some thing's missing here that doesn't make logical sense. So help me with this. Help me fill in. And so I think I got pretty good at telling the story. And often also being able to see things from an outside perspective. For example, one of my early jobs was writing up these graphics manuals for a workstation manager. And it was there that actually I first got the idea of creating some of my own books. Because it was clear that they didn't know what their users needed. You know I was there this is called Massachusetts Computer Corporation or Masscom. And I'd finished the job that I had been hired for and I had noticed that whenever any of us had problems with Unix, we went to this guy named Tom Dixera. You know we'd say hey, how do you do such and such and Tom would do some magic. And I said, "Well what are our customers going to do?" And they said, "Uh, we hadn't thought about that." You know because it was just like we were just Unix man pages and I said, "So are we going to write a book about all this root stuff, right?"

And so I went and interviewed Tom and this other guy Terry who were there assistant administrators and kind of wrote what ended up becoming the first ever Unix system administration manual. And I then made a deal with the company where I licensed it back from them and then proceeded to sell it to a number of other companies. Eventually I actually bought it back from one of those companies. There is a book that we published now called Essential System Administration by Eileen Frisch. And actually Eileen was a tech writer of a company called MultiFlow that had licensed my original book, rewrote it there. And then when she later came back and sort of flushed it out; it was a full book for us. But you know a lot of that---

Bob: So there is bits of you in Eileen's work.

Tim: Yes, yes, still bits there. Well it's pretty far down. Many, many years. It's like Troy.

Bob: So the theory in Unix world was that if you, the man pages, you know the online manual was the answer.

Tim: Yeah, in fact for a lot. I still believe the man pages a brilliant innovation in documentation. And you know there is a lot of focus in for example the world of Open Source on having access to source code. But if you look at Unix culture, an awful lot of the value is in the fact that the programs are small and atomic. And each one of them comes with it's own documentation.

You know so if you asked me would you rather have the windows source code or a man page for every DLL and component? You know which one would be more empowering to users and developers. It would probably be the man pages, right. Because then they would go, oh, I know what that piece does. I can take it out. I can get rid of it. I can replace it. But it's this big obscure lump. And where as Unix had this culture where all the pieces were contributed by other people. And they had this culture. You wrote a little bit of documentation for it.

And what we added as a company was the next layer up. Where we would say, gosh, this is a pretty big, complex program. You know all it has is a man page. What else could we add? Or maybe it was a research paper that was published by Bell Labs. Classic example of that was the paper on SED, the stream editor. So ___ showed how you could transform lines in Coleridge's Kubla Khan using this SED stream editor. And it really wasn't very helpful. And I discovered SED and regular expressions early on. And I just love that stuff. It was like wow, this needs a book. And so---

Bob: This needs a book.

Tim: Yeah, that's right. That was really how we got into it. We were tech writers. I came up with the idea that I wanted to develop products because I saw services businesses being a dead end long term. You know you just end up on the treadmill. And we had this idea that well our first idea was simply to retain the rights to manuals that we wrote. So I did a number of books for MassComp. I got the rights back from that. Some other clients and language manuals and the like. We were getting people asking for the same thing. I need a ___ manual or I need a C manual.

And we were retaining the rights, selling them to subsequent clients. We'd say, "Okay, yeah you want a manual." Normally you'd expect to spend this much in six months. Well you can spend 20 percent less and have it in six weeks. If you license our version, we discuss the mindset for you. And that it went that way for a while. But we also as a consulting company, we had downtime and we didn't want to sit on our hands. So we started saying wow, some of these Unix programs could use some better documentation. So then we started writing little books in our spare time.

And a lot of the original books were very small. The first books we published were 70, 80 pages. They were really about individual utilities. And that was really the idea that really eventually became our signature brand was this idea of the animals. And that was because they were about individual programs. And this one designer was like these Unix programs sound like weird animals. The other thing is a nice bit of history. This woman, Edie Freedman was not actually an employee. We didn't even know her at the time. Given our later you know association with Open Source software. It always struck me as sort of a wonderful piece of delightful irony that one of our most valuable assets that animal brand was actually contributed to us by someone we didn't know.

You know she was the housemaid of one of our writers. We had commissioned some designs for books that I just didn't like. Our writer went home over the weekend, was happening to talk about this and she the designer, Edie was a designer at Digital said, "Oh, it sounds like a fun project." And she just did seven covers for us over the weekend. Sent them in and they were pretty weird. They were the animals with the big eyes or the lumpish flesh. They were the weirder ones in the set. We put them on the wall and we sat with them for a week and we said, "What the hell."

Eventually Edie came to work for us and was our creative director for many, many years. She still works for us. She's now doing editing of books on graphics. But it's a wonderful piece of just luck, pure luck.

Bob: And you think the animals contributed to the success?

Tim: Oh, absolutely. Because first of all when we first started publishing books, we didn't know anything about publishing. And we had a lot of dumb luck as a result. So a good example on the front with the animals was that we well it was a series of things. We had these really pamphlets on individual utilities. They didn't have spines. They were stapled. First of all they were too small and then we also believed that our customers wanted them to be able to lie flat in the way that you could do with a skinny perfect bound book.

So that meant they got shelved in a separate section because they couldn't be put in with you know these other books. So they would be off in some corner somewhere in this face out display.

Bob: With the comic books?

Tim: Yeah, sort of, yeah. And so all of our competitors were like who are these guys? How come they are getting this whole display space? Everybody else was like paying money for displays whereas we just get it because that was the only way they could shelve the books. And the animals were weird and different and distinctive. And it's kind of like a lot of good brands. This is a slightly cult like element. There's a backstory, some people know it, other people don't. In a lot of ways the animals became iconic for programs.

It happened over time. In the beginning they were just kind of what is this weird thing? When we did a very popular series of books on the X Windows system. It was more aimed for corporate adoption. And we didn't put animals on them because you know they people in corporations would have turned up their noses at the animals. But over the next five, ten years, they became you know kind of something that's known and accepted in you know virtually any corporation.

Bob: Did you go back and put the animals on them then?

Tim: No, they, we never did it. The X books actually kind of died out as a line. We also we tried over the years to do to protect the brand, we actually often would we would have books that didn't seem like they fit the model. We would not put an animal on it. In fact we used to laugh. We hired a guy from Addison-Wesley, one of our competitors. And he said, "We always used to cheer whenever you didn't put an animal on one of your books." And I said, "yeah, that's why we're strongly branded and you're not. Because we know what an animal means."

So we would have books, I still remember Dale Doherty who is our first editor and I the first time we decided not to put an animal on a book it was a book on POSIX. And which was sort of this standardized version of UNIX. And it was just it was a good book, but it was kind of a dry reference book. And whereas the animal books that we'd always written had had this almost kind of I would say this joy of our liberation from the constraints of being hired technical writers. You know where we basically could tell people what was wrong with the program. You know just practical advice. And this book was a good book, but it was just a little boring. And we said it doesn't deserve an animal, so we just kind of did a text treatment. And it sold okay, but it was really, we really felt the animal meant something. That it was hands on, it was practical, it was the information from the trenches.

It was the kind of stuff that in our tech writing consulting days, we always wanted to put in the manuals. You know kind of where you go back to the environment that we started publishing in. You have to realize this was still a world where we would have clients who would say, "Oh, it's inappropriate to use the second person in a technical manual. Or you can't tell them about things that don't work." You know this was always a struggle. You know you'd say, "Wow, this feature doesn't work or it's buggy." And you couldn't say that. And so part of what was so exciting to us when we were able to do our own books is we could tell people "It's not you." You know, "it's really the program, it really doesn't work."

And I remember having that debate with authors. You know where they say I can't write this chapter yet because this feature doesn't work. And I go, "That's what you got to write." Because otherwise everybody is going to be struggling with the same problems. And of course the culture changed quite a bit. But I think for a long time there was this period where manuals were very dry. And they were just about the product, not about the conversation with the user.

A lot of what we try to do is to capture that feeling that we actually experienced as consultants. Where you go ask the guy at the next cube for advice and you know they'd say, "Here's what you do." And it was very practical, very hands on. And a lot of our early books, we just grew over time with user advice. We published our e-mail addresses and I still remember one of our signature books early on was a book on program called uucp, which was communications program for Unix.

And I think the first edition was about 80 pages and it grew over the next 10 editions which was about four or five years. So we were doing a new edition every six months based on user input. People would say, "You don't have any information on the such and such port contender. You know and here's how you configure it." And I just drop that right in. You know? And so the book grew organically in response to user feedback. And I really miss those days because we in some ways we've gotten bigger. And we still try to be very responsive to readers. But you know you move into a more traditional publishing process and we have been working to try to get back to that level of porosity if you'd like.

Bob: Porosity?

Tim: Yeah, poracity, where you just engage the market as you're developing a book.

Bob: It sounds like you started essentially in the Unix area though you mentioned digital and of course digital did more than just Unix. But you're best known these days I think for your relationship with Open Source. How did that come about?

Tim: Well, first of all I have to say that I really believe that Unix was largely Open Source and the Open Sourceness of it actually predates any of the awareness of Open Source. The licensing was very generous. And so for example back when I think Richard Stallman started the free software foundation in 1984. Well I had complete Unix source code in 1982, 1983. And so did everybody else I knew. Right, because it was on very liberal licensing terms and source was supplied. And it wasn't sort of a political issue at that point. But it was just what you took for granted.

And in fact if you look at Unix strictly with all the work that went on at Berkeley, many of the components were contributed by Universities. You know things that were core Unix utilities that we were using, VI or curses, the screen manipulation library. They all came out from Bill Joy at Berkeley. Then it'd come from AT & T and so there was already this sort of Open Source sharing culture. Everything that we later came to characterize as Open Source was there in early Unix. You know the difference was that because of the licensing terms, AT & T was able to come along and say, "Well, this is a very nice party, we think we'd like to own it now. So all of you, the rest of you please go away." And that was when you started to have the revolt that led to Open Source.

But all of the behavior was actually there before any of the licensing that we now come to associate with Open Source. And for that reason that I actually I've tended to feel the licensing issue is a massive red herring. And you know it's really about you put developers together with networks first of all. And you have architectures where and probably the biggest thing that drove Open Source in my mind was purely the architecture of Unix. And there were two different aspects to that architecture that were completely essential to the formation of an Open Source culture.

The first was there were all these different hardware architectures. And so if you didn't have source code, you couldn't distribute a program. There was a lot of sharing and freeware on the PC but or on the Mac or on the Osborne, but people would distribute binaries. But because Unix was a Babel of different hardware platforms, the only way you could distribute a program was as source.

So it was really an accident of the fragmentation of the Unix market. And the second aspect that was really critical was that Unix had this sort of idea of what an operating system was that was minimal. Where you had a small kernel, you had all these utilities and so somebody could write an additional utility that played by the rules that had been set forth. And it just worked. So there were kind of ideas that you wrote to standard out. You read from standard in that ASCII was the common exchange format for a lot of programs.

So it was really pretty easy to build programs that could just get plugged in and they just worked. So that was fairly essential. And then over on top of that we started to build a consciousness of the value of sharing. But so I think of my roots in Open Sources in that early Unix where I first became conscious of the values of Open Source was through the X Windows system. And in particular it was through pissing off a lot of people by taking the documentation that had been developed in MIT, released under the MIT license, which said basically it was like the ___ license. It said do whatever you want with this. And what we did was we said, "Wow, this is sort of a good start on the documentation set, let's take it in and prove it and enhance it." And we didn't keep it free because unlike the GPL, the Berkeley and the MIT licenses allow you not to do that.

And a whole lot of people were miffed. Bob Schiffler though, the guy who was the head of the X Consortium said to me. Said, I went and asked him, I said, "Gosh, should we, what should we do?" And he said, "You are doing exactly what we want people to do. Our goal was to create a platform that other people would innovate on. All these people are doing it in software, they are taking our work and incorporating it into their systems and this is exactly what I had in mind." And that was when I formed my core value about Open Source, which is respect the wishes of the creator of the software.

So if somebody for example because there are a lot of different goals. Somebody might say I really want to get my software out there as much as possible. Somebody would say I want people to build on top of this. Somebody else might say I have this political aim that this be used to advance the goals of making all software free as Stallman has done. And so we've tried to respect that as much as possible.

Bob: How? I mean what do you do differently in different instances?

Tim: Well you know a good example. We have published a number of books under the GBL and it's usually for us it's been around asking the author what they want. You know so and I guess it's that pragmatic I have a pragmatic approach. So a good example Richard Stallman for a long time argued that all of our books on free software ought to be free. Because the documentation that you know goes with a free program should be free also. And I actually I go, "That's a noble goal, but I think it's better to have some documentation than no documentation." And take for example sent mail. I went for six years through six authors and I finally found a guy who would do it but only for a very large advance. And he wants to maximize his revenue.

And that's his goal. And that's the only way I could get him to write this book. So are we better to have a no free book or are we better to have a non-free book that I have to pay somebody a lot of money to do. I think we are better to have the book than not. And so where as in other cases, somebody says "I would love to write this documentation for free. I want to advance the cause of free software." And therefore we then would say, "Okay, great, we'll put it out under the GPL." So our book on Samba is GPL'd or some of our early books on Linux were under the variations of the GPL.

But the other thing besides that sort of idea of just being practical in utility is also to figure out how do you maximize value? I wrote a paper once for it was sort of an online symposium that nature did about some scientific publishing. And I wrote a little piece about free software. And I really talked about a phrase that Larry Wall used. He said, "People say information wants to be free. Actually information wants to be valuable." And really Open Source and commercial software are both strategies for how somebody makes their work of their minds more valuable.

So some people say, "Wow, I can make this really valuable by starting a company and building up that way." And somebody else says, "Well I don't have the instincts for that or the interest in that. I could make my work really valuable by giving it away." And so to me Bill Gates and Larry Wall just had different strategies. They had the same goal. They wanted to create value. And I think that that's a really legitimate set of choices. And I think we as a culture benefit from the creation of value in both forms. And so a lot of what guides me is looking at how do I maximize the value of what we do? We have a limited amount of time here. And what we want to do is to make it count. We want to make things better. We want to make things new. We want to make things exciting. And so how do we do that?

And so I often look for sort of maximum utility in my discussions. Anyways so back to sort of free software. I didn't give it, so I was sort of having this shaping experience in the late '80's with Bob Shifler where I kind of became an officially an auto of the MIT Berkeley style of license more than of the GPL. And but I didn't think a lot more about it until early 1987. And that was when I really began my sort of actual Open Source activism. And the reason for that was actually a conversation with Carla Bayha who is still the computer book buyer borders. And she happened to tell me that the second edition of our book Programming Perl, which we published in '96. 1st edition was in 1991 was one of the top 100 books in any category in borders for all of 1996.

And it struck me when she said that. I thought wow and I don't think I saw a single story about Perl in Infoworld or PC Week or any of these magazines. And it really bugged me. And right about that same time, I'd just before that had hired Andrew Schulman who'd written a book called Undocumented DOS and Unauthorized Windows '95 and the like. And he was fond of sort of poking holes and sort of Microsoft stories. And at that time, Microsoft had been running a TV campaign about Active X. And it was all Activate the Internet with Active X. And Andrew pointed out that everything in the TV commercial was done with Perl. He had done some work on this except for this one little animated taxicab that went across the screen. And that was the Active X control.

And it just it really it totally cracked me up and it made me think actually of this great story from I'm a classics I studied classics in college. And this great story in the story of Jason and the Argonauts where they have been saved from certain death by Athena. And they see Apollo flying by in the distance. And the sailors all fall down and worship Apollo. And a feminist scholars have tended to sort of pick out this passage and a sexist kind of thing, but it just happens a lot. Where the real credit goes to somebody who just got the tinsel. And so I thought about that and I gosh, Perl is really important. So I decided to do a Perl conference to try to spread the word.

Bob: Was this your first conference?

Tim: That was our first conference. So that was in the summer of '97. And in so I was really focused on why PERL was so important. And I spent a lot of time thinking about that question. It's actually been a shaping thought that I've still. It's actually still influencing how I think. So the first public, well actually I guess I had given some earlier talks about Internet standards. But I gave a talk at this is before the Perl conference. But I kind of developed a talk called Hardware, Software and Infoware, which was about this idea that just as the IBM PC had somehow changed the rules of the computer industry, that somehow the Internet and Open Source was actually wasn't Open Source at that point. But something was happening now to change to this new kind of application.

And it was thinking about the importance of Perl on sites like Yahoo and the like. What is it that they are doing you know that makes a Yahoo and I coined this term Infoware. To say well hey, these guys they're, what do they do with Perl at Yahoo? One of my authors Jeffrey Friedl who wrote a book called Mastering Regular Expressions. I don't know what he does there now, but he used to basically write these massive PERL scripts to match up news stories to Ticker symbols. And it was like that whole idea of wow, there is this new kind of data oriented software. Where people are in the bowels of the beast using Perl and other programs to build these so dynamic databases. And that's very different from the kind of software you had in the PC era. And different enough that I tried coining this other term which never caught on.

But anyway I was, I went to this Linux conference in Germany. And I was giving that talk. And Eric Raymond who I never met was giving the Cathedral and the Bazaar. It may have been the first time he had given it. I heard it and thought it was great. And so I invited him to give that talk as a keynote also at our first Perl conference, which was a couple of months later.

So that was when I started really to think about it. So I did the obviously like a lot of other people I got all fired up by Eric's thinking about Open Source and the most general sense. Although again it was not yet called Open Source. And but then after we did that Perl conference through that fall I kept thinking about the fact that actually all of my best sellers had this characteristic that they were sort of off the radar. That nobody was really paying attention and so I decided to and I also realized through a series of conversations I was going to IETF meetings for example. I was also going to USENIX and I was going to various kinds of meetings with other communities. And I suddenly realized that I kind of knew these disjoints sets of people that I thought all of these things were connected. But other people didn't seem to know each other. So here I am in one context, here I know Larry Wall, but I wouldn't, all these guys who I knew from the free software world, they wouldn't be at an IATF. And so I thought, gee, I really ought to introduce all these guys to each other. Because they actually have something in common. A guy like Paul Dixey who wrote the DNS. And Larry Wall who wrote Perl or Linus Torvalds who wrote Linux. These guys all have something in common. And they ought to all be together and share war stories.

So I created an event that I originally called the freeware summit where I invited in about 20 people to brainstorm. And it was at that meeting that one of the issues that came up was the problem of the term free software. And then Eric Raymond at that meeting said, "Hey we were in another meeting a couple of weeks ago. And somebody proposed this idea, Open Source." Meanwhile actually Michael Tiemann of Cygnus was pushing the term Sourceware and we actually ended up having a vote. What we did was we said, we actually planned this event as a PR event. So we invited a lot of press after the day. So we have to go out and tell them something. What are we going to tell them? And so one of the things that we told them is we realize this is a problem. And we've all agreed, we took a vote and we agreed on this new name. We are all going to start using it.

Bob: And that was Open Source?

Tim: That was Open Source. And ---

Bob: Wow.

Tim: And we got up there and had these guys we felt we had about 10 or 12 of them up on stage. And we had a pretty good press attendance. New York Times, Wall Street Journal. A lot of tech publications. And I remember saying these guys all have something in common. I said, "They have no money but they've all created industry leading programs that totally with the strength of their ideas." And we just kind of started to go out and tell this story. And I remember it was one of those times when you could see over a period of a couple of weeks just how the consciousness changed. It was incredibly satisfying. So I periodically had it maybe 10 times where I'd just get in the middle of some tornado of some big idea. And then I'm on the phone with press for weeks on end. And this was one of those. And I still remember the riff I was using at the time was I would ask every reporter. I'd say, "Can you tell me what the five most mission critical programs are on the Internet?" And they would scratch their heads. And I would say, "Okay, well let me tell you what I think they are. I said the first one is a program called Bind, which you probably never heard of. And it's written by this longhaired programmer Paul Vixey who you never heard of. And it's free software. And second one is probably Apache which runs X percent of the web service. And third I'd probably pick Sendmail because it routes most of the Internet's e-mail." And then I went down the list, got to Perl and then Linux actually. And Linux was kind of on people's radar. But the rest of them weren't.

And the first, I still remember it like the first three or four or five days everybody I talked to it was like complete news to them. And by the second week, it was like everybody was like oh yeah. And it was just that incredible satisfaction of injecting this new idea into the newest sphere and having it catch. And that was the thing that I guess I would say more than anything else I was focused on was I felt like Linux was getting traction. And again I was really interested in Linux. We started publishing on Linux on the early '90's. And I think it was a great program but I was always a little perplexed in that people weren't drawing the story in the right way. They could understand Linux because it was an operating system. And they could kind of do this simple Syllogism and it's like Windows. It's another one of these face offs.

And I saw this much bigger picture. And so what I've spent a lot of my energy on was trying to get people to see these other things as important. And I knew from my book sales that they were important. GS & ___ was a bestseller. Everybody used it. Programming Perl was a bestseller and it was like, "Hey guys, pay attention. All these programs are important." So anyways, so that was I guess another thing that to just sort of bring up in that context. Actually kind of rolling back the clock to 1992. Is really where I first got the idea of activism as marketing. And I always like to give credit to the guy who taught it to me. This guy named Brian Irwin. He was, he had been the director of activism for the Sierra Club. He lived here in Santa Rosa, was tired of commuting to San Francisco and we managed to hire him as a PR guy.

And he came in immediately and said, "nobody cares about books, and they care about ideas and issues." And fortunately when we hired him was just when we published the first book on the Internet. It was called the Whole Internet Users Guide and Catalog. And Brian said, "We got to go out there and just push the Internet. Tell everybody how important it is." So we sent copies of the book to every member of Congress. We sent them to, we went there and we pitched I was used to going to Unix review and Unix World. And he was like, "No, we got to go see Time Magazine and the New York Times." And we organized this press trip. And I was like who is this weird guy coming in and shoving this book in my face.

Bob: And how many books did you sell?

Tim: We ended up selling over a million copies of that book.

Bob: There you go.

Tim: And of course it did in fact lead to so many things that were important. In fact one of the mosaic developers I believe claimed in a magazine interview once that their team, the NCSA learned about the web from piece of O'Reilly junk mail about the book. Actually that was one of the things that we got in the book at the last minute was a chapter on the World Wide Web. It wasn't actually written by the author of the book Ed Krol, it was written by Mike Loukides, his editor because Ed was sort of a slow writer. And Dale Doherty who had been our really one of our pioneering editors but also be very interested in an online publishing had discovered the web. Even though there were only about 200 websites at the time. Thought it was just the most amazing thing. And we had to get it into the book. And he arranged the meeting with me and Tim Berners-Lee in Boston and must have been the summer of '92 right before the book was going to press.

So we just got Mike to write this chapter at the last minute before it went to press. And that is where most people learn about the web and you give it this huge boost. The other thing that came out of that was that there was a catalog part of the book which had been we wanted to make free. It was again kind of going back to this idea of free software. We had the original document that we based the book on was something called the Hitchikers Guide to the Internet. Ed Crowl had written that at the University of Illinois. And made a free document. So we expanded it in the book. And we wanted to give something back. So we decided to give back this catalog part, which was modeled on the whole earth catalog. And it was just a list of all these resources on the net. So it was actually predates Yahoo or anything. It was really the first catalog.

Bob: Was that the global network navigator?

Tim: Well it became the global network navigator. So what happened was that I had this idea that people should try the Internet. So we actually had a project where we were going to put Internet kiosks into bookstores along with stacks of the books that people could try. And we actually ran a computer literacy bookshop but events kind of people started to get the Internet. We may have rolled out a couple of the bookshops as well.

And we had gotten, Dale had gotten involved in the web originally because we were interested in online publishing. And our X Windows system books in the late '80's were had been pretty widely adopted by workstation vendors. And they were all doing online documentation systems. You know HP had something called Answerbook, I'm sorry Sun had Answerbook. HP had Laser ROM. IBM had something Info Explorer or whatever. They all had these fancy names. And they all wanted our books in their format. And we thought that was a bad idea and kind of influenced by Bob Shifler in the X Windows system. What we thought was well what we want to do is we want to create one format and that went to something called the Davenport group that Dale started which created a document format called Docbook which was actually an SGML DDT at the time. It later became XML for technical documentation was actually developed around the structure of our X books.

And then we basically had found this guy who was a student as UC Berkeley, Pei Wei who'd written a browser called Viola where it really was a hyper card clone called Viola and one of the applications was a web browser. That's how we discovered the web, we were actually looking for a free bookreader to go with Docbook so that we could say to these vendors, here's our documentation, you guys can learn to read this format. If you don't want to read it, there's a free reader that goes with it. And but Pay basically when going back to the Internet, we had to put this catalog online and Pay with Web browser, said I could make a cool demo for that kiosk you want to build. And he built this really it was the first point and click interface to the Internet where you could just click on a link and you'd go to telnet server. You'd go to this link; you'd go to a web server. And it was this point click version of this whole Internet catalog. And I looked at it and I said, "Pay, that's not a demo, that's a product."

And it was actually InfoWorld that kind of gave me the business model for that. Because we were trying to figure out, Dale had this idea he wanted to do an online magazine and it was going to be a quarterly. And I said, "Gosh Dale I think the web is going to be the kind of thing that people have open on their desktops everyday. So we have to think about updating it all the time. And what's the business model going to be?" Anyway my copy of InfoWorld showed up and I thought, "Wait, there are these magazines that are ad supported that are free. We could do it like that." And that was where we came up with the idea of doing ads and we thought ads have never been acceptable on the Internet. But we thought this is really different because it's really on our site. People could come to it or not.

So we decided to launch in with advertising. The very first advertiser on the Internet was our lawyer at that point was Dan Appelman of Heller, Ehrman. They were the very first person to write a check for Internet advertising. And it was of course the web was so much in its infancy that the advertisements were actually hosted on GNN what we really and we had we always been idealists. You know our idea was this was the chance to recreate advertising. And so an ad was really a resource center. What we now think of as a website, we basically built a website for Heller Erman and hosted it inside GNN. And then we had kind of a yellow pages where you'd have a little button that you'd click on the link and it would take you into this sort of set of information about the company.

So in fact if you think about advertising on the Internet as all the popups and all the BS that's come since. We didn't create that, but what we created is the real advertising, which is this idea of you have a commercial site about your company about your products. So that's what we really pioneered. It was this huge foo foo; there was an article in the Wall Street Journal. You know the Internet is being hit by ad clutter. But we decided to go for it and obviously nothing has been the same since.

Bob: No, no. But you eventually sold the global network navigator?

Tim: That's right.

Bob: Why?

Tim: Well, I always wanted to keep the company private. I mean first of all we were out in the boonies completely homegrown company. Not very sophisticated. I think I've gotten a little bit more sophisticated over the years. But we weren't in Silicon Valley; we weren't part of Silicon Valley culture. We just didn't know what to do. And I read a book by Bill Davidow one of the founders of Mohr Davidow called Marketing High Technology or something like that. He made a really compelling point, which just stuck in my brain. And it was to dominate the market; you have to be more than half of the market and growing faster than the market as a whole.

I looked at this and I went oh. There is no way we can dominate the Internet because we don't have, we are a small company, we don't have the resources. I don't know how to do it. So that was when we decided to sell. And I could have if I had been more sophisticated; I would have gone and found a VC or whatever. But we met with a bunch of VC's and part of it was they looked at me and I went I was just this idealist. I was like I had all these ideas about wanting to sort of spread knowledge. And I was also; my original business plan for the company was interesting work for interesting people. It wasn't go out there and make a lot of money. And I had this whole idealistic story about why I was in business, what I was doing. And they looked at me and they went, I had this great conversation with Bill Janeway, of Warburg Pincus who took me aside after one of these meetings and said, "You don't want to do this. Don't take our money, you wouldn't like it." And actually Bill's been a friend ever since and he's now actually on my board. But yeah I think it was partly just lack of sophistication. I was kind of ---

Bob: So you had to make a choice that's helpful.

Tim: I had to make a choice. And I really felt that I wanted to be able to control my destiny. And another part of it was really having been a consultant around a lot of startups. When I was a tech writer originally and I'd watch these companies, they'd go from these idealistic high-energy places to one more boring big company. And I just didn't want to do that. And I had this idea that we could just do interesting things and make good things happen and grow organically. And in fact it has. The fact that other people have gone on and built billion dollar companies while we are still relatively small company. In some ways I made the wrong choice, but in other ways, I don't regret it. We've had a tremendous influence.

And the fact in our we did a management retreat I remember in probably '98 or '99. And the thing that stuck with me from it is this idea that somebody put, I think it was Brian Irwin put up on the board. Which was create more value than we capture as a company goal. And I think that we've been pretty successful at that. We've created way more value than we've captured.

Bob: That's interesting.

Tim: Versus there is a lot of scuzzy businesses' goal is to capture more value than they create. And there are a lot of wonderful businesses that create a huge amount of value and capture a lot more of it than we have. No question that we have been on the light end of the capture side. Relative to at least I think to the value that we manage to create.

Bob: Now you sold global network navigator to AOL. If you were so unsophisticated, how did you come up with that deal?

Tim: Well we had hired a woman names Lisa Gansky as general manager to work with Dale. Dale was the head of GNN, he was the guy who kind of got us into the online area in the first place. He was our VP for online. But he was not really interested in doing the business sides. So we gone out and tried to find someone. And Lisa was the one who basically Lisa's a great schmoozer, knew everybody and I don't know quite how she hooked up with AOL originally. We spent a lot of time with phone companies for example. In terms of trying to build the business for GNN we would do things like go meet with Pacific Bell and say, "Wouldn't this be amazing, this new web thing. Think about what it would be like if you could just have it offer it to your customers and they get it was part of their phone bill." And they were like, "yeah right." But they missed out that opportunity.

Bob: Bell heads.

Tim: Yeah, right. They just didn't get it. And so and we just saw that the web was taking off and that we didn't know how. We knew we'd eventually be marginalized unless we did something. So I think from that point of view it's probably the right thing. Now if we held out longer, maybe we would have been acquired for more money somewhere else. Or this was really one of the first transactions of the Internet era.

Bob: You know last week we talked to Brewster Kahle. And---

Tim: Yep, sold about the same time.

Bob: Sold WAIS to AOL about the same time they were buying GNN. So it looked like they were really putting together something that they then didn't leverage very well.

Tim: Yeah, I think what happened was that they were nervous about the Internet. And so they sort of assembled all these properties. And then they went actually it's okay, our existing service is fine. It's we don't really need to make that move. And then they kind of just didn't really focus on it. And then a few years later of course they regret it. I saw Steve Case maybe it must have been 2000 or something like that. The first words out of his mouth were "Oh, I still regret how we muffed the GNN opportunity." Because it really was the first web portal. It was Yahoo before there was Yahoo. And AOL they treat it as just kind of an off brand of their service rather than really continuing to embrace the web.

Bob: Well wasn't it more a combination of Yahoo and Netscape in essence?

Tim: Well, not really. Well we had a product called Internet in a Box that we did with Spry, that was a joint venture we did with a company called Spry that was sort of Internet access plus the content. But GNN was really it was just a web portal, a magazine. It was a lot of what Yahoo became. But the idea actually, the name came from actually a little interesting piece of history. Dale had found a book called the Navigator, which was a guide to navigating the rivers of early America. And so it was kind of this idea that people would need a guide to navigating this new land that was emerging.

But we actually probably our bigger mistake than selling to AOL was selling our AOL stock. What happened to us was as soon as we sold to AOL or relatively soon thereafter, we realized that they just didn't get it. So as soon as we could sell, we did. Not realizing that the boom was going to happen. If we had held onto the all the GNN stock to AOL's peak it would have been a billion dollar deal ultimately. But we sold. I mean again we did alright out of the deal so I can't complain. We did a lot of other interesting things. But we just went in there and one thing I still remember being in AOL headquarters and hearing the hissing of modems. Those people were dialing up all the time. Oh no, they really don't get it they are still in the dial up culture.

And I remember having all these arguments with them about the fact that people were eventually going to be online all the time. And no they just didn't see it, so.

Bob: Oh well.

Tim: Yeah, but that's.

Bob: Now at that time Microsoft didn't see it either.

Tim: That's right. In fact we went up and demoed GNN to them and they were incredibly dismissive. It was like I wasn't actually at that meeting. But Dale went up with a couple of people and felt pretty insulted. That they were like this is just stupid. I had one conversation with it used to be the CT of Nathan Myhrvold in '94 I think where he had this great talk at a PC forum. And it was all about. It was really about what Chris Anderson. Later Chris and the long tail. It was really about yeah he literally went up there and did a graph it was like the long tail. It was all these documents that are read by, there is a few documents that were read by millions of people. There were all these documents that are read by only one of two people. There is this really interesting region in the middle. And we have this mechanism for capturing that. It's called the Microsoft Network.

And I went up to him afterwards I said, "Nathan, this is a great talk but I think you're wrong. I think it's not going to be the Microsoft network, I think it's going to be the Internet." He was like "No way." So I always remember those early conversations with, it's certain satisfaction and seeing how it has developed a lot in the way that we foresaw it.

Bob: Well there are very few people who can say that they were smarter than Nathan Myhrvold.

Tim: I don't know if smarter in that particular case seemed from a little bit of a sideways perspective.

Bob: Now you have spent now a career dealing with writers and programmers essentially. How are they different characters, personalities?

Tim: You know I think the thing that I would say more than anything else about our customers is that they are creators. You know what I mean? A lot of we've actually launched a magazine recently called Make in the term that we've been introduced as makers, people who make stuff. And there is people who make stuff in the physical world. There is people who make stuff with words. There is people who make stuff with programs. And I really believe that that whole creative culture, people didn't realize how creative programming is. And anybody who's done it of course knows that not only is it creative, but it's incredibly absorbing. Getting it's I know there has been a lot of debate even among programmers about Paul Graham's essay, Hackers and Painters. It certainly matches my experience. That it's a creative act and there is a lot in common with that kind of you have this image of what you are trying to create. And it's like the old story of Michelangelo chipping the image finding the image in the stone. You have this idea of the way you want that program to work and you are just going to chip away until you find it.

And this bug is in the way. I find that creative streak I think often leads in a programmers to be good predictors of where culture as a whole is going to go. And that is where I think I've tried over the years to in some ways use my customers as a filter or a predictor of where technology as a whole is going to go. Or where the world as a whole is going to go. So really the early adopters of the Internet were all scientists and programmers. But they did things that were sort of the native way that you work in that culture. So they shared a certain way, they communicated a certain way. And so I was able to say boy as more people get online, these kinds of behaviors will spread. And the kinds of things that happened when there were a few hundred thousand people online would become very different when there were millions of people online and different kinds of people.

And so I have the talk I give sometimes called Watching the Alfa Geeks which has that premise that you often that people are very comfortable with technology who can kind of bend it to their will. They often will build these incredibly complicated things. And people will say the geeks aren't good. They don't understand the average consumer. And they don't have to. But they can tell you where the technology wants to go.

And so two examples that I use a lot are one was web services. We gave our first talk on web services what is now called web services in 1997 at our very first Perl conference. Because it was obvious to us that this is where the Internet was going. Because all the programmers we knew were treating websites as data sources. They were writing these Perl programs to scrape data from some remote site. We for example, we had all these programs that scraped Amazon for competitive information about our, about other people's books and pricing and all this kind of stuff. Why would we go mouse around on for hours at a time and type it into a spreadsheet? You get a programmer to do that for you.

So the programmers were building all these sort of half ass ways of using these sites in unauthorized ways. And we said wow, this will get easier eventually. People will figure out how to do it right. And sure enough bit-by-bit that has turned out to be the paradigm for how people have started to think about the platform today. And that's the whole focus of the conference that we have now, web 2.0. It was their implicit in our very first conference nine years ago because the hackers were doing it. And we were saying okay, eventually the entrepreneurs are going to come along and say we are going to make this easier for ordinary people. And then it will just become part of the platform. Part of the fabric of how people think.

Bob: So what's happening right now that will be that way five years from now?

Tim: Well the thing is there is a whole lot of things that I look at. Probably maybe one of the most striking is I think the rise of virtual worlds. In a place like second life is in some ways the purest example of it's not necessarily the most widespread. Obviously there are games that have 100, 100 million players where second life has 50 or 60,000. But what's interesting about second life is that it's not a game. It's just a place. And it has events and it's just the people at Linden Lab, the people that build it. They just think that they are Real Estate developers. And they've created this unimproved land and they've started to build an economy that is very different than the normal gaming, you know people basically are renting space. And then they are improving it.

If you improve land in second life, you own it. You can sell it to someone else. You've built something on it. And there is starting to be real life events. You know people have book signings in second life. They have people go give lectures there. And I think that's going to really take off. Where the boundaries between our physical life and our virtual life will become increasingly blurry. We've seen this in science fiction. And periodically start-ups get ahead of themselves. I remember when there was a whole virtual reality craze. It was right after Neil Stephenson's what's the book, I can't remember the title (Snow Crash). But the one with black sun in it. And there was a company called Black Sun that was going to be a VR company. And everybody kind of gets ahead of themselves. But fact is if you look out the long through line, that's certainly one area that I believe will increasingly take off.

Clearly there is this whole other area that we are calling Web 2.0. This idea of the Internet as platform. Also look at things like Wikipedia and how collaborative development and really radical approaches to collaborative development are going to take hold in more and more fields.

Bob: But you were doing that before when you took, when you were revising the book every, at least twice a year?

Tim: That's right.

Bob: You were the Wikipedia.

Tim: That's right, we were doing it ourselves. And again that's exactly the example. Using whatever crude tools, at that point our e-mail address were UUCP and it was all dial-up but that early community was still have that sharing, this idea of people building on each others work. And I think that's a natural approach to information sharing in a network era. I think we are going to see more and more of that.

Bob: But are books going away?

Tim: You know it's an interesting question. I think that ---

Bob: The answer is no.

Tim: I think the answer is it depends. And when I think about books, I think about I think a lot of people lump them into a single category because of their form factor. And I think you have to think about them according to the job they do. Because there is certain kinds of books that are I think are almost definitely going to go away. Other types are going to flourish. So for example, if you create a reference book, you are very much in competition with the web, right. And the web does a good job of reference. And if you sort of write fantasy novels, you are in competition with a game like Everquest or World of Warcraft. You know they do the same job.

And I think there may always be people who like books. But I think the market will potentially shrink quite a bit. And so what we've tried to think about are how are books going to change. And so for example we've never been very interested in stand alone e books. The reason is because if the job, one of the jobs that we do is referenced, we have to ask ourselves, how do you make reference better online? And ways that you make reference better are one, searchability; two is having it be up to date. Three is having it be potentially be collaborative with links across difference documents. And these are all things that the web is really good at.

So that was the origin of our safari service, which is we now have 3,000 plus books online as a joint venture with Pearson. We also have contact from Microsoft Press because it is people want to use it for answers. And they want to search across more than one book. So moving towards the large body of content on the one hand, but we also are in the business of training. Teaching people new things. And they're, it's not quite clear that people have cracked that nut. We are starting to see there are certain kinds of tasks that you could train pretty easily with for example screencasting is good if you want to show somebody how to do an application.

And so I think that there are many types of application oriented books that might go away. There is other kinds where the book is still the best way to teach somebody these complex concepts or whatever. And or even just having the additional band with of here's the open book. Here's the screen. I don't have to so there is different ways that a book is useful. But I actually think that books are going to change and quite a bit over the next decade.

Bob: Yeah, but we'll still have books.

Tim: Oh yeah, I'm a huge collector of books. I personally own six or seven thousand books, so I and I certainly don't want to see them go away.

Bob: Good, we're done.

Tim: Alright.

Bob: This program was made possible by a grant from the corporation for public broadcasting. This is PBS.

Be more connected. Help support your PBS station and projects like NerdTV...pledge online.

© 2005 NerdTV LLC and PBS. Offered under the Creative Commons license.

Privacy Policy | Terms of Use | Published on September 6, 2005