developerWorks: You're listening to developerWorks interviews where we feature conversations with technical luminaries and thought leaders from a variety of disciplines on topics of interest to technology professionals. I'm your host, Scott Laningham. Our guest today is Tim O'Reilly, founder of O'Reilly Media Inc., and an important supporter of the open source movement. Tim describes O'Reilly Media as a technology transfer company changing the world by spreading the knowledge of innovators. Tim, thanks so much for making time for this. We really appreciate it.
O'Reilly: Glad to be here.
developerWorks: Now, first of all — I'm really interested in life patterns of people, especially important successful people, and I'm wondering what does the typical Tim O'Reilly day look like, or weekday anyway, if you don't mind me asking?
O'Reilly: OK. Well, actually — the first part of the answer to your question is weekends don't look too much different from the weekdays. Maybe the workday ends a little earlier. I usually get up first thing, 6:30-7, check my e-mail. Usually have, you know, another 40 or 50 unread messages, some of them personal, some of them mailing lists I follow. And I check blogs that I follow, and often I will spend an hour or so actually writing an entry for my blog if I have time.
developerWorks: You're not a 4:30 in the morning person then. That's good.
O'Reilly: No, not a 4:30 in the morning. I usually get started about 7. If I have a few days at home, I'll often spend a few hours writing up blog posts, which I'll then schedule out for the week because when I'm traveling, I often don't get to them at all. And so, speaking of travel, a typical day might actually be somewhere else, in which case it's a very different schedule.
developerWorks: You've accomplished so much with your book publishing and your Web presence, and the O'Reilly conferences. Is this where you expected to be, or has it caught you at all by surprise?
O'Reilly: Well, you know — I guess when I started my company, which was back in the late '70s, we were actually a technical writing consulting company. And I certainly didn't imagine becoming a publisher, conference producer, Internet pundit, any of the things that happened in that sense. But when you look backwards at your life, you see a pattern. And the phrase you used earlier, changing the world by spreading the knowledge of innovators, this was something that I at some point looked back and said, "Oh, that's what we do."
And you know, it really started by being a company doing documentation for high-tech startups, and I noticed that a lot, in those days, a lot of what was offered as computer manuals really didn't match the experience of people in the industry. I would say, "How do you do such and such?" And they'd say, "Oh, go talk to Joe." You know? And you'd go talk to Joe, and he'd kind of show you some things. And there was this very interactive peer-to-peer kind of relationship. And I said, yet here we're offering our users manuals that say, you know, the user should do this. And people in those days even frowned on using the second person in a manual. And as I was doing documentation, I was finding that I'd say, "Well, this feature doesn't work." And I'd try to explain a workaround. They'd say, "Oh, you can't tell the user that our product doesn't work."
And so I started writing my own books partly as a way of getting around that so I could capture the authentic voices of the people who had learned how to do something, and they'd figured out what worked and what didn't. And so the user who was stuck wouldn't be there saying, "What am I doing wrong?" They might say, "Oh, it really doesn't work; nobody else can make this work, either." And I think that that kind of perspective was where we started, even in our documentation consulting business. And then, as we sort of developed a fairly unique style of doing documentation, we realized that a lot of customers were asking for the same book. It was the earlier — the rise of UNIX® and other standard software.
What we started doing was rewriting our contracts with vendors so we could reuse the material with other companies. And then from there, it was a small jump to start actually publishing the material as books. So that actually happened a little bit accidentally. We had a big contract that got canceled and we said, "What are we going to do with all our people?" So we ended up writing a few of our first books really just as a flier and took them down to a trade show. And before long, we realized that being a publisher was actually more fun than being a contract writing house. And so we started, really, our publishing career.
Probably the next big breakthrough was in 1992, when we happened to publish a book called The Whole Internet Users Guide and Catalog. It went on to sell a about a million copies and was kind of the book from which a lot of people learned about the Internet. And among other things, it was the first book to cover the World Wide Web. There were only about 200 Web sites, but we discovered this technology. We thought it was just fabulous. So we were evangelizing it.
But about this time, I hired a guy who had been the director of activism for the Sierra Club to do PR for me. He said, "Look — people don't care about books. They care about ideas." So rather than trying to promote our book, what we did was we went out there and used our book to promote the Internet. And that was really the beginning of an activism kind of marketing that we've become known for. So I became, you know, the prophet of Internet commercialization, went and spoke to congressional aides, and we did the first study of the advertising potential of the Internet in early 1994. We actually created the first commercial Web site — it was called The Global Network Navigator — in 1993. It was the first advertising-supported site, with a special dispensation from the National Science Foundation because there was an acceptable-use policy that said no commercial activity.
You know, all these kinds of things were really just around this idea of spreading the word of this new technology. So obviously, the Internet took off. We actually ended up selling GNN, as we called it, to AOL. It was really one of the first big .... It was actually the first content-oriented transaction of the dot-com era. That was in June of 1995. And then AOL managed to box it up, and that was the end of that.
But we continued to do our publishing and technology activism. But what probably the next big breakthrough we were known for really came in 1997. And that was when I published the second edition of a book called Programming Perl — Perl, of course, being a free software programming language. And when I published the second edition in 1996, it was a huge success. And in fact, the buyer at Borders, the bookstore chain, told me early in 1997 that Programming Perl had been one of their top 100 books in any category for all of 1996. And I thought, "Wow — how odd. Nobody is talking about Perl." And then I realized that they weren't talking about any of my bestsellers. And that there was this whole set of technology that nobody was talking about because it didn't have big companies behind it. It was all developed by individuals. And we had books on many of these technologies, and the grassroots adoption was huge, but there was no recognition. You know, you could read the computer trade press and there were, few ... people were starting to recognize LinuxÂ® and maybe a little bit about it. But nobody was saying, "Wow — the Internet runs on, you know, a program called BIND, the Berkeley Internet Name Demon that runs the domain name system, and this is maintained by a single long-haired programmer that lives in Redwood City."
developerWorks: It was very grassroots awareness at the time.
O'Reilly: Right. So what we did is we decided we were going to try to get that word out. So we organized a meeting that we originally called the Freeware Summit where we brought all these guys together .... Because that was the other thing: We realized they didn't all necessarily know each other. So we brought them all together, and in the course of that discussion, one of the issues that came up was that the name "free software" was ambiguous and confusing. And at that meeting, Eric Raymond proposed that we try this new name that came up at a brainstorming meeting he had been at a few weeks before: "open source." We actually took a vote and everybody decided to use it. And that was obviously a historic moment because we had a press release ... I mean a press conference organized for the end of this summit, and we had enough credibility with the press that we had some people from Forbes and Fortune and The New York Times and Wall Street Journal were there.
So I got up there with all these guys up on stage, many of them at that point unknown, and I said, you know, "These guys are creating a revolution in the software industry with nothing other than the strength of their ideas and this new model. And we have a name for it. It's open source." And you know, within three weeks, Linus Torvalds was on the cover of Forbes, and there were stories in all the major media, and that was really in a lot of ways the PR launch of the open source movement. Now the movement itself had existed long before that. But it was that kind of — for us, it was recognizing the pattern that was happening in doing some storytelling about it for the industry.
And then probably the next big thing that we got a lot of attention for was in 2003, after the dot-com bust, you know, everybody was down in the dumps about technology. And we were still very bullish. So we were trying to figure out how to tell the story of, you know, there was a lot of life in the Internet revolution yet. So we decided to launch a conference called Web 2.0 to highlight that there was a new revolution all these new companies coming along, and they all had something in common. And I wrote a paper called "What is Web 2.0," and described the pattern of how the software industry was changing, and I guess it caught on. Web 2.0 has become such a buzzword that we're all kind of sick of it at this point.
developerWorks: You've certainly been part of the authoring of some of the most powerful phrases of the day, that's for sure. You mentioned Web 2.0 there. And I'm wondering what other arenas are you looking at for O'Reilly Media, especially with Web 2.0 opening up that field of possibilities.
O'Reilly: I guess there's a couple things. First of all, our basic methodology, as we've developed it over the years sort of through trial and error, is that we find interesting people who are innovating from the edge. And then we just watch and see what they do. So, for example, we have an event called foo camp, which is — I can't remember who came up with this term; it might have been Steve Levy of Newsweek — called it the wiki of conferences because ... we invite these guys together with no program, and on the Friday night, they introduce themselves and then there's a bunch of big whiteboards with space for talks, and they put up the talks that they want to give. And we watch that and we say, "Wow — what are they wanting to talk about?" And so there's this sort of emergent data that tells us, "Oh, wow — there's this movement in these communities towards new technology."
And so, for example, one of the things that came out of foo camp for us was how much hackers are starting to work with hardware again and not just with software. And that led us to do some publishing, you know, sort of on the hardware hacking and eventually to launch a magazine called Make, which has been a huge success, which is really ... and in fact, coined the term "maker" to kind of describe this new class of person who is not content with sort of the stuff that they can, buy but they want to remake it. And then we kind of sit there and we look at something like that, and we say, "Well, what are the trends that are driving this?" And in the case of this whole maker revolution, there's really three or four overlapping trends that drive this.
One is there's an awful lot of disposable technology today. So that, you know, you're on your 10th cell phone. What do you do? You just throw out the old ones? There's new uses for them. Unfortunately, terrorists have found some of those uses, as timers and detonators for improvising explosive devices. But there are much more interesting applications of used technology. For example, one of the projects in Make was a guy used an old VCR to build an automatic cat feeder for when he's away. It gives you kind of an example, if you go out there, you find these people doing these crazy kind of things. But what we've noticed is that often mixed in with the crazy things that are just, you know, one-offs and fun, there are ones that are illuminating. And I want to come back to that in a minute because I wanted to hit the rest of the trends that were driving this maker revolution.
The next one is that the tools of CAD/CAM and design are becoming far more ubiquitous. So, for example, with SketchUp and Google Earth, you can actually design things, or Second Life, you can design in 3-D.
The third trend that I think is really, really important is this fab revolution, so to speak, that Neil Gershenfeld talks about, where we really are moving in the direction of desktop manufacturing. I mean, you can now get a laser printer or water jet or 3-D printer for about the price you could buy a typesetter back in the beginning of the desktop publishing revolution.
And then the fourth trend I think is people are in some ways getting sick of the consumer culture, and they're looking for new ways to find fulfillment. And all these other things are making it possible for people to make things and not just to make software. You know, we're seeing an increasing number of particularly these hardware hacker types. They're turning their interest to alternative energy. So that the whole idea that global warming is a serious issue has penetrated the hacker consciousness, and people are saying, "Well, we can do something about that."
So you know, there's a lot of innovation happening. This kind of goes back to this idea that I was going to expand a little bit. Just this idea that hackers show us something about the possible shape of the future. And so, for example, when Wi-Fi came out, you know, the developers of the technology saw it as local area networking technology for offices and the like. Hackers immediately started building homebrew antennas up on rooftops and beaming Wi-Fi down to their local coffee shop. So you could actually see that that was a coming trend from this hacker activity. It didn't mean everybody would be building homebrew antennas; it meant this would become sort of a ubiquitous wireless mesh, which is the direction that we're going in.
And so similarly, back on this power theme, we see these little things bubbling up. So, for example, Lee Felsenstein had this project that he called the pedal-powered Internet in Thailand where he was setting up Internet access for people where they were generating the power with a bicycle — you know, with a stationary bicycle because there was no electricity there. Now, does that mean everybody is going to be moving that way? No. But it was sort of like a first sign of a power-related project. And now for example, the $100 laptop's getting a lot of attention, and power supply is pulley-generated. You know, in fact, that was designed by somebody who was at our foo camp, so we saw that coming.
So, we're seeing those innovations sort of bubble up. The other areas we're starting to see a lot of people thinking about AI a little bit, which was such a deadly word for a long time. But machine learning is becoming hot again. And of course, that comes back around to the whole Web 2.0 idea, which we've sort of skirted. I've spent a lot of time talking about in the last few years, and that really is sort of an understanding of what are the big changes that result from the Internet becoming the central platform.
developerWorks: Let me ask you a question going into that, if you would, because on that subject of Web 2.0, I read that excellent piece of yours, "What is Web 2.0?" on the O'Reilly site. I think you may have known this, but I should say, we had Tim Berners-Lee on back in August, and I asked him about a common summary explanation that floats out around there on, about the phrase "Web 2.0." And actually, I wish I'd read that article before I'd asked the question. But what I said to him was, a common explanation out there is Web 1.0 was about connecting computers and making information available and Web 2.0 is more about connecting people and facilitating new kinds of collaboration."
O'Reilly: And then Tim said, what? Are you kidding? Web 1 was totally about connecting people. It was a groupware product.
O'Reilly: Well, I would have given you the same answer because I think it's unfortunate that a lot of people latched on to the term Web 2.0, and they've used it to mean, oh, it's rich experiences with Ajax. It's mashups. And for me, it's a much broader concept than that. And in fact it is, Tim Berners-Lee is absolutely right; it is no different from Web 1.0. Web 1.0 was the first Web 2.0 technology. It was actually probably Web 1.5 that went wrong, you know, which was all these people with business models where they were trying to make it look like television. Broadcast, you know, one to many.
developerWorks: So Web 2.0 is post- ...
O'Reilly: Web 2.0 really is a fulfillment of Tim's original vision and the vision of all the Internet pioneers because what it really is is an understanding of what it really means to build network applications. And the only reason we put the 2.0 on it was really to indicate that it's a new phase in the evolution of the business model. It's not really of the technology because what happened was everybody got all excited about it, and they went wrong. And then it crashed, and then we went back and some people got it right. And the people who got it right, we were trying to highlight tell a story about what made it right.
For me, the very heart of Web 2.0 as I define it — I've since boiled that paper down to a one- or two-sentence definition — is it's the fact that the Internet is becoming the platform, and because the Internet is a platform, it's possible to build true network applications. And a true network application is one that uses network effects to get better the more people use it. And if you think about it and you apply that to a whole lot of different success stories on the Internet, you start to see how that idea of network effects leading to applications that get better the more people use them is at the heart of every single success story.
developerWorks: All the way back to hypertext.
O'Reilly: Right, exactly. The original World Wide Web. I mean, it's a system that gets better the more people connect.
O'Reilly: And then you think about Google. Their breakthrough in search was page rank, which was not brute-force analysis of documents, right? It was figuring out something about the link behavior, what people were doing and making the system smarter by analyzing those links. Now, also, I mean the fact that a Web search engine grows simply by crawling. Every time you or I make a link, we're feeding Google. We're contributing to it in the same way somebody might contribute to an open source project by sending in a bug. We just don't know that we're doing it, right? It's implicit in the structure of the Web and that's why the Web is, in fact, an incredibly powerful system — has an intrinsic, what I call architecture of participation; people pursuing their own self-interests make the system better.
And P-to-P applications, even though they're not, this is where it's more than just the Web. P-to-P applications are also Web 2.0. I mean, when Shawn Fanning said when you download a song, you also serve songs, he was realizing something that you could do that made the network more powerful. And even though there were copyright issues there, the ideas of P-to-P are still alive and well. And Skype is the latest incarnation that shows just how powerful that idea is. Here they have 5 million simultaneous users, and I believe they own only 12 servers, right, because they either own 12 or 5 million depending on how you count.
You know, you look at other network effects businesses. Amazon — it's an e commerce site, but what made it different from Pets.com, which was a dot-com flame-out, was that Amazon really understood that you needed to get your users to make your product better. And you know, you look at the level of annotation. Amazon compared to any other e-commerce site works so much harder at getting people to participate. They've got more than 10 million user reviews. On every page, there are probably half dozen or maybe a dozen invitations to participate. And I think that that idea — that you build systems that get better — is something you can do perhaps on a stand-alone platform. But it's the natural way to make Web applications.
You know, the creators of a Web application can do something that you can't do with a stand-alone application, or at least not without adding on a lot of networking. You can learn from your users. You can try a new feature. And do the users use it? No — we must have done something wrong, or they didn't like it. Let's try something else. So it's almost like, I mean, MicrosoftÂ® has a great name. They call it, you know, WindowsÂ® Live because the idea is the software is live. It's not just something you create as an artifact and then it's over. And I think that sort of crept back into traditional desktop software, you know. For example, you think about Microsoft's Dr. Watson. Your system crashes, it sends a bug report automatically. That's something where the network is being used to help the software get better the more people use it, right? So to that extent, Windows has become a Web application.
But the point is some people understand this very purely, and they're making it the heart of their business model. Eric Schmidt at Google has a saying that he likes to remind people internally when they're doing product planning. He says don't fight the Internet. And it's that sensibility that I think is a key to the Web 2.0 era. And it really is simply, you know, when you have a new kind of technology revolution, what you have is the possibility for some people to understand the new rules of business better than others. And so, for example, you could say, looking at the personal computer revolution, you say "What's different about it?" A personal computer is just a computer, right? In fact, computers, the personal computers of today are as powerful as or more powerful than mainframes of 1980. But that would be missing the point because there was a huge shift when the computers became personal. And first, it was a shift to standardized hardware. And that led to commoditized hardware, where it became cheaper and cheaper, and the margin went out of hardware and that made this huge change in the business where it drove value to software. And that was something that IBMÂ® did not understand that Microsoft did understand. And Microsoft, as a result, became this hugely successful company while IBM struggled for many years. Now, IBM later was able to regain their feet.
I think in a similar way what's happened now with the Internet. We've entered an era where software has really become standardized. And yes, there is proprietary software still, and Microsoft still is a very profitable company. But the frontier of software development is networked applications. And networked applications by their nature, require standardization because both ends have to speak the same language. And so, unless you can get to sort of pre-emptive marketshare early, you're going to have to be standards-compliant. And in the case — for example, of the Web — Microsoft was able to dominate the browser, but they were never able to dominate both the browser and the server because Apache had a free and open source implementation of the Web server standard, and because that maintained dominant marketshare, they have to more or less comply. And so the software tends to become standardized, which has the same effect that standardized hardware had. It's become commoditized. So I think we're moving into this world where the value is moving out of the traditional software ecosystem.
But Clayton Christensen, the Harvard Business School professor, has this great concept. He calls it the law of conservation of attractive profits. Profits don't just disappear; they move somewhere else. And a lot of the story of Web 2.0 is the story of "Where did the money go?" And as software becomes commoditized through standards and open source, the value has moved into this new advertising-driven online ecosystem.
developerWorks: So a summary thought around all of that — I mean, Web 2.0 really is about getting out of that big pothole that was the dot-com bust and continuing on down the same road that Web 1.0 was already, right?
O'Reilly: Absolutely. And it's really — there's a wonderful line from the poet Walt Stevens that I love. He refers to searching the possible for its possibleness. And you know, we're really in that stage with the Internet, where the Internet has become widely deployed and we're now saying "Wow — what can we do if we really understand the power of the Internet?" And I think we're not there yet, all the way there yet. And all this innovation is still exploring what gets better as it becomes networked.
An example that I give people all the time is cell phone. You think about cell phone software, and you think about the way, for example, your phone remembers who you call. Then why hasn't anybody built a Web 2.0 cell phone address book? That would be one where your cell phone providers' call center gets turned into your social networking application. The fact is because this is your real social network all the people you call and message with. And why do I have to tell my phone to remember that stuff? Why isn't that just something that my phone automatically looks up that information from, you know, the big database in the sky? Which exists; it just hasn't been connected because nobody has figured out yet on the carriers that there's a new possibility that you could build a better kind of networked application. So someone will do that, and then we'll have this advance.
So I think it's just really, if you look at these applications and the evolution of the Web, it's really — or the evolution of the Internet — it's really people understanding, "Oh, I could do that better." You know, MP3 dot-com said, "Let's put all the songs in one place," and this college kid who grew up in the age of the Internet said, "Why would you want to do that? The songs are already where they are. All we have to do is point to them." And that was a further, a deeper understanding of what it meant to have the Internet be the platform because a lot of people are still stuck on the idea that the Internet is sort of this add-on to the PC. I think now everybody knows that it's front and center. But you know, in that transitional phase, that's where a lot of people got stuck.
developerWorks: If we see it more as an ideal or as a world view even if you have it, instead of a group of technologies that keeps changing or evolving, then is it really unnecessary or inappropriate even to ask questions like, "Well, what's the adoption cycle for Web 2.0?"
O'Reilly: Well not really because Web 2.0 isn't a single thing. You know, it's kind of like saying ... well, I guess you could say it in the broadest sense, in the same way you could say "What's the adoption cycle for the personal computer?" Right? If you do it at that level, it's certainly appropriate. You can say, "OK — more and more applications are moving to Web 2.0." There's an adoption rate, for example, by which applications that used to be formerly isolated desktop apps are now collaborative online applications. I think it would be hard to measure, but it's certainly possible. But it's not like the equivalent of the adoption of say Windows Vista vs. Windows XP because it's not a technology; it's really a technology movement. And the fact is that, you know, the boundaries between a personal computer and a mini-computer or mainframe just aren't meaningful anymore.
developerWorks: Where are we, do you think just with getting to a place where we really feel like Web 2.0 is where we're at?
O'Reilly: I would say that we are about at — if you look at the personal computer, it's defined actually by its applications more than by ... you know, like nobody remembers anymore and says, "Oh yes, the real milestones were the introduction of the, you know, the XT or the 386." I guess they are milestones. But you often look at the software milestones. So there was the Visicalc era; people realized it was a new application that you could do on personal computers, and that was the killer app that really drove the business side of the personal computer revolution, and word processing was sort of secondary.
But then there was LotusÂ® 1-2-3. Somebody came along and figured out how to do it better. They made it graphical, they made it ... they did the integrated, you know, word processing, spreadsheet, graphics. And then along came the platform stage where Microsoft was able to leverage their operating system. There was a move to Windows. And as part of that transition, Lotus lost and Microsoft applications kind of took over. And then following that was probably the real ascendancy of Windows over the next five or six years where it was all over, the period where people were saying, "Will OS/2 be a real contender? Will UNIX dethrone Windows?" And you know, there were all those kinds of discussions, and then there was kind of a game-over period.
So if you look at that kind of timeline, I think you would say that what I'm kind of referring to as Web 1.0 is the Visicalc era; I think we're kind of at the Lotus 1-2-3 era, and we're just starting to enter the shake-out period where some of the platform players are starting to emerge and flex their muscles. But it's not clear who is going to be the winner yet. And you know, if I follow that sort of line out, I think there really will be a platform winner. And not just a single winner, but certainly likely to be some. There's definitely going to be some accretion of power in a few places. And then what happens is the innovation goes elsewhere because once there's a winner, they tend to raise the barriers to innovation.
developerWorks: Want to predict any platform winners?
O'Reilly: I mean, it really is too early to call. But I mean Google is obviously a front-runner. I think Amazon is an interesting player because they're kind of doing some explicit platform plays with S3 and EC2, that are getting some traction, whereas Google's platform plays are all about their own applications. But still, Google has such a powerful economic engine that's sort of tied to the heart of their business. And it's pretty clear that sort of search in some ways is the, you know, owning search is becoming the equivalent of owning the GUI in the personal computer era.
developerWorks: Yes, an amazing company.
O'Reilly: Yes. And I was talking with Jonathan Miller last night, ex-CEO of AOL, he was kind of saying, look, you know, if you look at, one way to look at it is that, you know, AOL and Yahoo! and even Microsoft's online efforts, MSN, are really all e-mail companies. That's what really drove them. And I'm not sure it's entirely true of Yahoo! because Yahoo! is also a Web content company. But e-mail is, at least from Jonathan's point of view, a big part of the traffic. And he said, look, Google is really the first company that turned search into a sort of a platform business. Now everybody wants to be Google. It's kind of a shame because, hey, if Yahoo! wasn't in the search business and if they were just evaluated as a content, you know, everybody would be really happy. They'd say, "This is a great business." But because they got into the same business Google, everybody said they're not Google.
developerWorks: Software developers are a huge part of our audience on developerWorks. What do you think are the most important Web 2.0 skills that developers and students need to be building right now?
O'Reilly: Well, I think probably the first and most important thing that distinguishes Web 2.0 developers from traditional software developers is this kind of live software sense. I had this great conversation with Mark Lucovsky, he was the original designer of Microsoft's Hailstorm, you might remember, a senior guy at Microsoft. He's now at Google. And we had had an e-mail exchange about something that he was doing. And I said, "Oh — so, cool, can I blog it?" And he said, no, no, no — I haven't finished it yet; I'm going to finish it tonight. You can blog it tomorrow. And then he wrote, how cool is it that I can finish this software tonight and push it to the world tomorrow? You know, when I was at Microsoft, I'd finish software and I'd have to wait for three years for it to show up on the CD.
And you know, that creates a whole new set of dynamics in the relationship between a software developer and his audience. I mean open source is always, one of the key principles is treat your users as co-developers. And in the Web 2.0 world, that's really possible because you try something. Your users try it. They give you feedback — either explicit feedback or implicit by what they do and what they don't do. And then you change it. So there's a whole software development modality there, which is really tough for people, for example, who have grown up in the era of sort of waterfall planning and scheduled software releases. And I don't think that goes away. I'm sure there are applications that still require that. But there's a kind of dynamic live interaction. And a set of tools that support that that are critical. And so scripting languages — Perl, Python, Ruby — are much more optimized for that kind of programming.
developerWorks: Would you say that the real-time nature of open source development kind of unleashes some elements of creativity that were shuttered in that older model?
O'Reilly: Absolutely. Absolutely because you can throw up that first plausible promise and get feedback on it. And I think the tools, I mean I think Joe Kraus describes the difference between Jot Spot and Excite. He says, you know, it costs us millions and millions of dollars to launch Excite; cost me $200,000 to launch Jot because the tools are so much quicker. I think part of that also is we've made it possible for less-sophisticated programmers to build applications.
And I think that that creates, unleashes a lot of creativity because it's bringing new people. You look at the whole Adobe ecosystem with rich Internet applications and some of those things that are going on there where they're bringing designers and other people into the programming world. And I think a related skill I think is Internet-enabled collaboration. I think open source is really about collaboration, and there's a set of tools for collaboration. That's why, for example, you know back in 1999, the company that I launched as part of my open source sort of exploration was a company called CollabNet with Brian Bell with the Apache project. We realized that the right thing to do wasn't to commercialize the Apache product. There were a bunch of other people who tried that and failed. It was to commercialize the Apache process. And CollabNet has been an extraordinarily successful company by basically developing tools to enable online software collaboration.
developerWorks: What do you see coming next? What do you see beyond Web 2.0? We hear about the semantic Web, and we have even have heard the term Web 3 thrown out there. But what do you see?
O'Reilly: Well, I think, first of all, this idea of Web 2.0, because it's not, it really is the idea of searching out, you know, the true potential of networked applications. It's not going to be over for a long time. Whether the buzzword lasts, I'm not sure that there's a fundamental technology change that's going to happen from here on out. We're just going to get better and better at understanding what makes a good networked application. But I still think there's a few directions that I find very stimulating, and one is I want to see the emergence of a lot of very powerful data businesses because so many of these networked applications get their competitive advantage from their data, not just from their software APIs.
And there's new business models that go with that. One that I like to point to is a company in the U.K. that's now in the insurance business, and they offer pay-as-you-drive insurance. You know, in the old days, you set insurance rates based on where somebody lives. And they're saying no, no — if you have a GPS in your car, you can sign up for our program where we'll set your rates based on how far, how fast, and where you drive. Right? A whole new approach to remaking that industry based on real-time data collection from network devices. So if you think about that, you'll see that actually the next phase, if you like, of this Web 2.0 and maybe you call it Web 2.5 or Web 3, is that we get beyond applications that get better the more people use them as a result of people typing on keyboards or phones or devices, and the devices are just automatically reporting in. So it's kind of the sensor revolution, and everything that goes with that, I think is going to be profound and significant. With a GPS just as a particular class of sensor. Our devices are going to be participating in this network.
And then the other thing I think is really, really interesting is this idea that is embodied by Second Life and SketchUp plus Google Earth, and also by massively multi-player games, the idea the virtual worlds are really becoming new hubs for social interaction for commerce, for prototyping. And what's more, this is spilling over into the real world, whether it's through people creating virtual goods and selling them on eBay .... I mean again, going back to this theme of hackers showing us the shape of the future, these people who have started businesses where you can get an action figure made from your World of Warcraft Avatar, manufacturing things out of these spaces, to futures that you can imagine where your architect says you want to see the design of your new house, come to my office in Second Life and we'll give you a walk-through.
developerWorks: Yes, amazing. I've had you for long time, Tim. Any closing thoughts for the developer audience? It sounds like the future looks pretty bright, really.
O'Reilly: I think so. I mean, I think it is important to realize the future will not be like the past. And acquiring new skills is really important. I think that data skills are important. Hal Varian, who's a UC Berkeley professor who's done a lot of consulting for Google once said to me, SQL is the new HTML. I mean, you've got to be able to handle data these days. I think that getting out there and having fun is critical. A lot of, if you look at the history of technology, an awful lot of the innovations come from people just playing. And we're now in this phase where everybody thinks, "Oh, there's a lot of money in Web 2.0," but they forget that a lot of Web 2.0 success stories were started in a period when there wasn't any money — when people were just doing it because it was cool.
So I just kind of want to remind developers not to get hung up on trying to, you know, jump on the ... you know, to create the next big thing because you're probably too late. I mean maybe if you have a really great idea. Maybe there's still lots of opportunities, but the next big thing is actually going to come from passion, not from greed. I mean when Larry and Sergey started Google, everybody thought that search was dead. You know, nobody was going to make money in search. I guess I would just say, don't go to all the obvious places, just find something you're passionate about, try to make it happen, try to make the world a better place by your inventions. And if you do that, eventually you'll figure out how to make it work.
developerWorks: Great thoughts. Great closing thoughts. This has been great fun, Tim. Thanks for doing this.
O'Reilly: You're very welcome. Bye-bye.
developerWorks: Our guest again has been Tim O'Reilly, founder of O'Reilly Media, which is a technology transfer company that Tim describes as, "changing the world by spreading the knowledge of innovators." I'm Scott Laningham. Talk to you next time.
Scott Laningham, host of developerWorks podcasts, was previously editor of developerWorks newsletters. Prior to IBM, he was an award-winning reporter and director for news programming featured on Public Radio International, a freelance writer for the American Communications Foundation and CBS Radio, and a songwriter/musician.