Linux Weekly News Contact us

Open Source: the Basis for Web Applications The german magazine Computerwoche published an interview with Open Source advocate Tim O'Reilly. The original german version of the interview can be found on their Webserver. Thanks to Computerwoche for granting permission to run this translation. CW: What associations do you have when you hear "Open Source"? What most people will associate with Open Source is Linux but one of the things I have worked for, is to make people aware that OS (open source) is bigger than Linux and that it has a bigger influence. So, just as one can say that the internet is a network of networks, I think the OS community is a community of communities. I personally believe that for network communities of people "sharing ideas" is the key element. It is very clear that the low barriers to doing OS development, for example the financial barriers, have really enabled a lot of people to create new and innovative products. And those products, which I call infoware, are the future. CW: But is it necessary to promote Open Source as you do? Yes and no. For example, if you listened to the press, you would believe that everybody is using ActiveX for building their websites. But I have been at a web publisher conference where it was asked, "who is using ActiveX?" Nobody raised their hand. But when they asked, "who is using Perl", all hands went up. So OS is doing the work and all the publishing and marketing departments are getting the credit. CW: And that is one reason for founding the Open Source Initiative (OSI). Are you a board member of the OSI? Not directly. But I speak with Eric Raymond, the president of the OSI every couple of weeks. I urged Eric to be more involved with the board. The board of OSI was basically Eric and a couple of his friends and I have urged him to try to make it more representative of the actions of the OS communities. I am not sure that a lot of the leaders of OS communities were really thinking OSI matters all that much. I think that it would be good if OSI did matter - it was useful to these various communities. CW: The OSI is dealing mostly with marketing, but not with consulting. Part of the problem is that Eric is not really a business person. He is a provocateur and he is doing a good job of that. But somebody needs to create an organization which is effective... CW: ... and that needs money ... ... and that needs money, yes. When Eric sometimes goes and says to companies, "you just do OS and all will be well", I don't know if he knows what he is talking about. OS has some big problems. CW: What are the problems of Open-Source? Patents from commercial development? That is a major issue. CW: Is there anything going on in the Open Source scene? At these summits we have held, we are actually trying to create an organisation like the IETF. The IETF could be very interesting for legal issues not just standardization. People can work in the open source style on these inner/other issues, not just purely on software development as they currently do in the IETF. What needs to be done is to get people working on a problem and build a community of people who care about it. CW: At this Open-Source-Summit you mentioned, how did people from the corporations react to the invitation to the summit? Very positively. All joined the summit, including a representative of the white house. People from the corporations got it. Especially the guys from IBM, they asked very good questions. The guys from IBM traveled around and asked questions at various OS groups around the world. They do a very good job. Same thing with HP. CW: Larry Wall, the founder of Perl, is financed by your company. Which other Open Source activities do you push? We push several Open-Source-Activities. Larry has time to take care of further development of Perl. Besides Larry we temporarily hired Brian Behlendorf from the Apache Group. Brian is working on a specific project, which we hope to turn into a company. It is about refining the open source toolset. In addition we organize conferences and open source summits like the one in the beginning of March in San Jose. It is for bringing together the leaders of these various communities, because many times they do not know each other. And last but not least we publish a lot of books dealing with open source and open source products. CW: The open source scene has different camps: There are fundamentalists like Richard Stallman, pragmatists like Eric Raymond, non-political people like Linus Torvalds and so on. What would you say is your position? I would say pragmatist. To clarify, I like people who are curious, I like developers and I have no problem if a proprietary system competes with an OS system. But it is better to cooperate than to compete. So I would try to persuade Microsoft that OS is a good thing for them. What was for example the technology which caused people to upgrade from Windows 95 to Windows 98? It was the Internet. And Internet technology is based on open-source-software. CW: But Microsoft is not known as an inventer of key technologies and therefore competition on a technical level must be difficult for them. Yes, that is true. But alongside the success of open-source there will be proprietary software in the future. CW: In your Open Source book "Open Source, voices of a revolution" you said: "The point is that open-source software doesn't need to beat Microsoft at its own game. Instead it is changing the nature of the game." Could you explain this a little bit further? Well, a lot of it has to do with this idea of infoware building a new platform. The user is not only interested in databases, spreadsheets or operating systems. People are buying computers for different reasons and they are doing different things with them. And most of these things don't have to do with the traditional desktop productivity applications. They have a lot more to do with the transfer of information. Do you use the same kind of tools? No, you don't. What I see, is that many of the big websites are based on open source: Yahoo, Amazon... Two friends of mine bought computers simply so that they could buy books from Amazon and use Yahoo. CW: But that means that Microsoft hasn't got it. They think it's still the desktop, the PC. I think they will get it. I think they are a smart company. To beat Microsoft at the desktop is not the point. You see, the real goal is to beat the platform that the next generation's applications are going to be built on. And that is truly a difficult task: FreeBSD, Linux, Apache, Perl - these are the platforms that people are using to build the Yahoos and Dbases. Commercial proprietary software is starting to fight back in that area. Microsoft made a lot of progress with what they did after pushing ActiveX, which was sort of the old model. Then they realized their mistake, and now they are pushing ASP. They are starting to get it. So my message is more to the open-source community than it is to Microsoft: If the open-source community doesn't get it, they are going to end up fighting the old battle trying to win on the desktop. But you know, who cares about the desktop? The web is the platform. What you want to be is like "Intel inside" - you want to be like "open source inside" for the next generation. CW: As you know, many commercial companies port their applications to platforms such as Linux. How will industry react to the open source phenomenon in the future, how will open source change these companies? The OS Model isn't necessarily going to be a problem for companies. When somebody buys a product from IBM, what they are really buying is IBM standing behind the product. Who cares how it was developed? CW: But IBM doesn't support Linux directly, they support it via RedHat. But you will see more and more support directly by the companies. And what I think is very interesting is that the companies are really trying to figure out not just how to benefit from OS - they are really trying to see how they can help. At a conference in New Orleans, there was a question, "how many people think Microsoft should be broken up or should be sued by the justice deptartment?" 90 percent of the people were hostile to Microsoft. So I think that a lot of the companies are getting behind OS just because they are tired of being controlled by Microsoft and they would like to see that stopped. CW: How are open source communities organized in the field of software development? Quite differently. Take for example someone like Linus Torvalds and Linux. Linus will say, "This is bogus" and "go away". On the other hand the apache group has a voting system: any of the members of the core group has a vote. Before somebody can put in a major new feature they must get 3 members to vote for it but it can vetod by only one vote. That's kind of interesting.  CW: Can you imagine publishing books in an "Open Books" style? We have a number of open books. But first of all I am a publisher and I believe, that the rights to a book belong to its author not to the publisher. And if Richard Stallman tried to tell me to publish my books as open source, I would say "don't talk to me, talk to the authors." And then if the authors came to me and said, "Okay, we want to publish the book as open source", then I'd ask them, "Do you want to make a gift to the community? If you do, then that is fine, but you make less money." Making money is, in my opinion, one of the most important motivations for writing books and it helps to maintain the book as a service. Writing open source software is different. Richard Stallman believes that the users of software have all rights to that software. It's really the question of, what is communism and what is capitalism? It's true, you know, Richard's campaign started when companies were acting in ways that were disadvantageous to the users. He plays an important role as someone who fights for the rights of those users. But at the same time he's going too far: he takes the rights away from the software developers and I believe that what we really need to do is to find ways that benefit them as well. CW: You are facing fundamental differences in writing software and writing a book. Yes. One of the fundamental motivations when people write software is solving their own problems. And if that is your motivation, then giving it away, gives you both. Larry Wall wrote Perl, because he was trying to do some things and it was easier for him to write Perl than it was to do it any other way. Giving it away does not cost him anything yet gives him something back, because other people work on his problem as well. So there really is a transaction of value. Although Eric Raymond in the "Cathedral and the Bazaar" talks about reputation being the reason why people will write projects, - but mostly, if you look at Sendmail: Eric Allman wrote it originally to solve a problem that he had. Writing books is different. You do not write books for yourself, you always write a book for the community. CW: But looking at the software this is not always true. Think about KDE or GNOME. Neither have been started to solve a private problem. Yes. There are arguments to be made that those projects may not work as OS projects. So sometimes it is not to solve a problem, sometimes it is just to hack the hell of it. But once you start doing things in order to make money and when you start doing things for other people you need to ask, "what is the transaction of value? Am I doing this for reputation?" You do OS if you can make it work.

Linux ® is a registered trademark of Linus Torvalds