Tim O'Reilly Reflects on 10 Years of OSCON

By James Turner
July 22, 2008

Or, you may download the file.

It's been 10 years since O'Reilly held the first OSCON. At the latest edition of O'Reilly's open source convention, Tim O'Reilly sat down with O'Reilly News to talk about the anniversary. He also reflected on how open source has changed in that period, whether Web 2.0 (a term he helped coin) has met his expectations, and how the nature of technical book publishing has changed.


James Turner: James Turner with O'Reilly News here live at OSCON 2008 with the man himself, Tim O'Reilly.

Tim O'Reilly: Hi.

JT: So this is the 10th anniversary of OSCON.

TO: Yeah; well I kind of think it was the 12th anniversary because really it all began with the Perl Conference in 1997. But yeah; it is the 10th anniversary for the whole shebang. [Laughs]

JT: So as far as we're reflecting a little bit, what is the--I mean obviously the Open Source has become a much bigger thing in the last 10 years, but looking back what--what's the arc been like?

TO: Well you know actually this is funny because actually I'm going to kind of recap this a little bit in my morning keynote tomorrow. If you think about the origins of Open Source it's kind of--there were a whole lot of people doing something just for fun; they were doing it because it was cool and then it got discovered, . We discovered first with books, and literally I started the Perl Conference because Programming Perl-Second Edition was one of the top-selling books in Borders in any category in '96 and there was no mention of it in the computer trade press. And I said, we got to do some marketing for Perl. So I thought of the Perl Conference really as a way to raise the profile of Perl, and then I immediately came back and said oh my God.

All my best-sellers are like that . So it was the fact that this stuff was below the radar and we started trying to put it on the radar. And first with this summit that was first called The Freeware Summit¸ and there we realized there was a name problem. And Eric Raymond brought in the name Open Source, which had been invented by Christine Peterson a couple weeks before at another meeting. We had a vote, and everybody said okay; we're going to start using it. You know it wasn't a foregone conclusion. I mean Michael Tiemann was advocating for Sourceware and there were various other names but we took it--we--what we did was we agreed that we would use one name from then on and we went out and told the story. We had a press conference.

And then of course we launched the Convention to kind of tell the story to the world, but we were the rebel army. And you all remember the--kind of the rebel army feeling of those days. You know actually in my talk I show a picture of the OS-Wars tee-shirt of Atlanta when it showcased in '99, which was the same year we started OSCON, OSCON proper. And you think about Eric Raymond's tee-shirt at the time, with the Gandhi quote, at first they laugh at you then they fight you then you win, and you look at where we are today and by most measures of what were counted a success at the time, we won . Open source --Microsoft has an Open Source program.

JT: Yeah; I was going to say that I don't think 10 years ago having Microsoft as what--as a Diamond Sponsor this year or--?

TO: Yeah. [Laughs]

JT: You know you just wouldn't--it wouldn't have gone together.

TO: That's right; but they've been a sponsor for a while and at first it was very controversial. We had people who said I won't come if Microsoft is a sponsor and now it's just kind of yeah; it's part of the dialogue. It's moving on. You know people like Gartner, I remember in 1998 trying to persuade Gartner to cover Open Source. And they said why? Our customers aren't asking for it. And I said what kind of research company doesn't do any research until the customers already know about it? [Laughs] And they said well, you know; and of course now Gartner has been running Open Source Summits for --. So yeah; Open Source is getting adopted in the enterprise. Although it's still actually, I say I have some results from our latest Open Source Research Report. You know it--the adoption is much, much higher in technology companies in the Fortune 500 say than it is in non-technology companies. But it's catching on in the non-tech companies. But at the same time, the problem that I think caught the Open Source community unaware and I felt like a voice crying in the wilderness for many years because I felt all along that there was really a paradigm shift in the air, which I eventually wrote up in the paper, the Open Source Paradigm Shift.

It just said the Open Source community is not realizing that they're changing the world and that the world is changing in a direction that we weren't expecting and that is of course what's come to be known as Web 2.0; the idea that the biggest application was on this server-side building these server-based applications that were in the cloud, that were starting to use the internet as a platform, and on that platform, yeah you leveraged all this commodity economics of Open Source software, but you built proprietary services, that are backed by proprietary data.

And I remember having a debate with Richard Stallman about this in 1999, the same year we started OSCON, so that's been almost 10-years ago, and at a conference in Berlin and I was using the example of Google, which was still relatively new to the scene. And I said Richard if you had Google source code you still couldn't do Google because it's data; it's machines; it's business processes. And he said it doesn't matter because Google isn't on my computer. And of course now 10-years later it's pretty obvious to everyone, even the FSF with the autonomous group that just got started--it's a working group there that this is a real problem and that we have to rethink the fundamental premises of Open Source and how do they work in an era of cloud computing? How do they work when the software you use isn't on your computer anymore; it's on the cloud?

JT: Right; the value of Google is that everything is tied together and that-- GMail and Calendar.

TO: That's right.

JT: And all the other--and YouTube and all this stuff is all tied together. I wouldn't want it all on my computer because you lose the value.

TO: That's right; well and you also couldn't--. Yeah; so exactly, so there's this fundamental shift that we created and it's a lot like what happened, I have to say, if you look at the history of the PC, the explosion of personal computing was driven by the commodity economics of hardware being standardized. And so a lot of my thinking about all that was really driven by looking at the history of the PC and going oh, it's just going to happen to software. And then there's this idea of Clayton Christianson; he said you have the law of conservation of attracted profits. You commoditize something; people figure out a way to make something else valuable and exclusive and it's human nature, it's life, and I guess what I've wanted the Open Source community to do is to get ahead of that curve and to accelerate the process by which we start thinking about open data, open Web services, what does freedom mean in a context where it's no longer about having source code and to a program that you can modify and run on your own personal computer because more--it ain't the personal computer here anymore?

And fortunately there's a lot of fantastic Open Source that's starting to happen, you know. You look at Hadoop so duplicating--. So Google; you look at Eucalyptus duplicating EC2. People are sort of taking up the challenge and saying you know we can start to build some Open Source code that you can use. Jesse Vincent's Prophet, trying to figure out hey, could we actually have P2P synchronization in the cloud rather than centralized synchronization in the cloud? Not to mention all the fabulous Open Source software for scaling and managing large sites, so I think there's a lot of innovation happening and in some ways I guess I would say that it is the first Open Source Convention where I haven't felt like I was sort of trying to get the message to people who weren't hearing it. I mean sometimes--there's this great passage in the Bible where Jeremiah, the Prophet, has been out preaching--not that I want to compare myself to Jeremiah, but he's been out preaching and nobody is paying any attention. So he starts preaching to the ground and say oh earth--earth, hear the word of the Lord. And I'm kind of like, oh my God; for years I felt like--I could see this thing coming at us like an oncoming train. And nobody seemed to care, but this year it's like I'm going God; it's bubbling out all over. People are tackling the problem; they're solving pieces of it; there's innovation .

You look at even things like--and big companies have been getting behind it, . Google kind of going oh my God; you know we're about to be displaced by the closed phone-eco system, so we've got to make an open-phone eco-system. You know so here's a company at the top of its game realizing that that open really matters to them. And it's as if Microsoft had embraced Linux in 1995. This is fantastic, because Google is saying wow; the Web for all the parts that are closed about it, there's still this open architecture, this open participatory shared culture, and they're saying oh my God. If we don't work to preserve that on the phone, that whole culture is going to be challenged and taking proactive steps.

You know you look at--so I guess what I'm saying, the amount of Open Source innovation that's kind of starting to line up around technology as it is rather than around technology as it was, makes me more optimistic about Open Source this year than I've been in a long time.

JT: You mentioned Web 2.0 and obviously you're well-known for having coined the term.

TO: Yeah; strictly speaking Dale Dougherty coined the term but he used it--

JT: Or you popularized it.

TO: --he used it to describe a set of things I had been talking about for a long time. [Laughs]

JT: Right; do--if you look back at what you thought it was going to be like when you first started using the term and what it's kind of evolved into today is it more than you thought it would be? Is it--are there things about it that disappoint you?

TO: I think what--I guess what I would say is that Web 2.0 is pretty much what I thought it was. It's not what other people took from the term; a lot of people at--for a lot of people Web 2.0 came to mean social media, it came to mean ad-supported business models, lightweight development, and for me it was always about the idea that the network was becoming the platform and that when the network was the platform, the rules of business would be different. And so I kind of have been sort of frustrated to see people both get it and not get it and so I was just saying for example, Gartner was saying that Web 2.0 is sort of in the trough of disillusionment in the hype--in their hype-cycle whereas cloud computing is approaching the hype-peak. And I go wait a minute; don't people get that cloud computing is just another name for one aspect of Web 2.0? You know it's like --because again for me, Web 2.0 was just a convenient name to sum up the idea of the internet as platform, which includes things like the programmable Web which is not the same as cloud computing, which is really this idea of hey somebody is going to build a cloud and you're going to host your applications on it. And we haven't yet built the distributed cloud that would be where that gets really interesting, . It's still hey; here's my cloud; here's your cloud. And there's going to be a lot of competition and you're going to build on--

JT: Right.

TO: --you're going to build on Amazon, you're going to build on --.

JT: Right; I kind of wonder if things like BOINC are going--are kind of one view of what distributed cloud computing could look like, although that's a very limited--.

TO: Yeah; it is interesting and BOINC if you go back, again this is a trip down memory lane for me because before we called it Web 2.0 I was calling it the Internet Operating System in--back in 2001 when we did our--we had our P2P and Web Services Conference. And in that first P2P Conference a lot of people were really puzzled because we had, we were featuring internet file-sharing, we were featuring Web service, and we were featuring David Anderson, SETI@home and people go what do you mean? What does P2P have to do with SETI@home and to me it was always about hey these are both internet platform applications. You know how do you think differently when you've realized the network is the platform? And I have to say Distributed Computation has been much slower to take off than certainly I would have expected at the time; there was a wave of start-ups for example including one that I invested in.

JT: Well it's a hard problem to think about--

TO: Yeah.

JT: --it's not generally applicable in a lot of--.

TO: Yeah; although the technology is kind of coming together in an interesting way because now with Multicore you're having to think about parallelism even for--

JT: Right.

TO: -- traditional computing, so I think if we really start to--and actually some really interesting work out of HP, the whole Memristor Technology potentially leading to a kind of a whole new paradigm for even how you would build the hardware that would then build--in some sense could even be almost like sort of a hardware architecture, at least as I was hearing Russ Daniels--was bending my ear about this. He's a CTO for cloud at HP; there's sort of a possibility of building a much more organic kind of computing where the hardware learns like the software does. And really and again--which is just intrinsically parallel.

JT: Just to finish off because I know you've got places you've go to be.

TO: Yeah.

JT: One other thing you've gotten to watch the arc of is that O'Reilly really started out as the first, you can almost say geek-publisher; it wasn't Addison Wesley or someone doing computer textbooks; it was a very--focused towards the programmer trying to solve problems. Obviously that--you don't--we don't own the niche anymore 100-percent; do you think it's a good thing that it's really--that there's a flourishing of technical publishers now?

TO: You know I guess I would say that the--first of all, I mean if you look at the arc of our business, when we started publishing about Unix in the mid-'80s nobody was really--there were a few other books on Unix but we were certainly the first to publish the drill-down. There were a few general books but people were shocked when we published a book, say VI, and I remember it must have been at--I don't know maybe around 1990 being at a conference, a publishing conference, and I would throw-up some of our --

JT: Titles?

TO: --some of our titles, . And it might have been even the early '90s, and I would say how many copies of this book do you think we sold? You know and people would say oh 5,000; whatever-- Learning VI at that time was something like 130,000 and people were just like oh my God; they were shocked that there was a demand for these niche titles. So then it was this explosion where everybody kind of got into the game. And--but now really since about 2001, the entire market for computer books has shrunk and it's not really competition from other book publishers that's been the challenge for O'Reilly; it's been the competition for online sources of information.

If you look at the difference in what we publish today versus what we published say five or six or seven years ago, we--our core titles are of much higher proportion of reference or sort of kind of tutorial reference, books that had a combined --you think back on our X books, we literally had--here's the tutorial volume and here's the reference volume. Yeah; reference is almost entirely online; a lot of tutorial is online and the books that are our-- our best-sellers today are books that are hard to put online. So the Head First books which are very, they're graphical, they're interactive, and we haven't figured out how to do a good online experience, so that the--so the deep-teaching books are among our best-sellers. The--or things that are much more fun-oriented; when we did The Hacks books or Make Magazine, this is technology as entertainment because technology as a reference, hey you just look it up online.

JT: Right; and Make is like you want to have it in print because you're going to be in the garage underneath the thing--?

TO: Or because it's just sort of the richer experience that you have. I mean it's true across magazines in general; they haven't suffered the same way newspapers have because there is kind of an experience of the rich graphics, the rich experience of that form. So we're in a very different kind of publishing environment. I mean O'Reilly's publishing business is still the biggest part of our business but for example, you noticed if you compare us to 10-years ago when we did our first couple of conferences, now conference is a major part of our business because A, one of the things we can do to advance our mission which we say to ourselves is changing the world by spreading the knowledge of innovators is by bringing people together.

And that's a really powerful way to do it. It's really powerful sometimes to capture a book, but there's also things like this where you're just going out and you're sitting down with innovators and catching some video and putting it up online. So this is not done ; I guess the thing that all of us have to remember when we think about our business is--whatever our business is--is what do we really do? You know when I talk to publishers I try to remind them; I say are you in the business of putting ink on paper and then binding that paper and shipping? No; you're in the business of doing something. What is it? If you're a Science Fiction publisher say, your competition isn't other book publishers; it's games like World of War-Craft right; it's movies; it's --. So you're in the business of creating an imaginative world and Harry Potter and World of War-Craft are really part of the same genre in a funny way .

And so we look at O'Reilly's business and we--there's three jobs that we do. One is we teach and so we've tried to expand the ways that we teach; two, we provide reference and so for example when we built Safari Books Online we said how could we make our books more useful online as a source of reference, (a) make them always available; (b) make them searchable, across the whole corpus--that's a better product for reference. And then the third we do is entertain and so you look at more products like Make that are really geek-entertainment. And so by understanding the job as opposed to the form factor we've been able to add additional features. Who would have thought 10-years ago that we would put on an event like Maker Fair which drew 65,000 people to a family--? It's like an old-style county fair except with robots right. [Laughs]

JT: You have to do one on the East Coast by the way because my son so wants to go.

TO: Yeah; we will definitely--we're planning that and we may do a couple, so that's a--. We haven't announced it yet but it's definitely in the works.

JT: All right; well Tim thank you very much for taking time--

TO: All right.

JT: --out of your schedule.

TO: All right; thank you. All right; take care now.

JT: Tim O'Reilly was talking to us here at OSCON 2008.

You might also be interested in:

Popular Topics


Or, visit our complete archives.

Recommended for You

Got a Question?