In one sentence: What is Open Source?
Open source software is software which everyone has the freedom to use for any purpose, to study, to modify and to share. If you want to extend it beyond software, you will have to find ways of defining what it is to use, study, modify and share other things.
Could you give examples for other things "open source"?
For instance, if you look at a textbook, the freedom to use would mean the freedom to use it for any purpose, which includes commercial as well as non-commercial, to use it for teaching as well as an art project.
The freedom to study – well, to study is not the same problem for many other things as it is for software. With software, it means to be able to see the source code behind the program that's running. But when it's a textbook, using and studying can be more or less the same thing: you read to use and you read to study as well.
Freedom to modify, now that's interesting because you need to have the right to modify the textbook. Unlike software, anyone can make changes to the textbook in terms of practically doing so. It's more the law that prevents you from doing that. So if it's a textbook under a licence such as the creative commons licence, that says, "you are legally allowed to make changes to this", then you have the freedom to modify the textbook. And the freedom to share is of course being able to distribute the textbook to anyone else.
If you take producers of open source software: What drives them to do this, to give away things for free? What are their motives?
For software developers, the most common motivation is learning and developing new skills. Because free software developers learn by doing and most developers are actually not writing a lot of code. They are writing small amounts and they're making modifications rather than writing new things. So for them, it's learning. They are studying something, they're modifying it, and then they give it back. Other people comment on it, criticize it, and they learn more.
But for others who are creating big things, there are lots of different reasons, which include publicity, the need to get a job, solving a problem that's not possible to solve in another way. And there are lots of companies which release free software as well, and they have perfectly commercial motives; it's often better for your business if you release software for free and charge money for providing services than trying to charge for the software itself.
It also means that you can use software that is produced by others for free. So you create free software which is built upon other free software. That also helps to reduce the need for a "big" motive.
In the late 1990s, you developed an economic model describing why it is actually beneficial for people to participate in open software projects...
I call it the "cooking pot model". There reason for that is that it's named after a hypothetical tribal cooking pot: If I have a fish and you have potatoes, we both realise that we get more value if we both put the fish and the potatoes into a common cooking pot and make a stew.
The reason this doesn't happen that much in a real tribal cooking pot is that we still have to decide after that how to divide the stew. Did you put in more? Did I put in more? If you always put in more but we always take out equally, for instance, that seems unfair.
All those problems go away when you look at information and knowledge goods, such as software on the Internet or even textbooks. Because the entire pot can be copied for everyone. We don't have to divide it: I get the whole pot, and you get the whole pot. And that's what allows everyone to feel that they get more out of it than they're giving. I only put in fish, but I got out fish and potatoes, and you only put in potatoes, but you got potatoes and fish out of it as well. And you got exactly as much in fish and potatoes as I did.
Why should I as a consumer choose Open Source products if I have proprietary products working quite fine for me right now?
As a pure consumer, you're only interested in the product; there is no such thing as choosing between open software or proprietary software. You're only choosing between one product and another. Just as you might choose an Apple Macintosh or a Windows PC, you might choose OpenOffice.
One of the main reasons for many end users of course is the price. While the word "free" there refers to freedom, the fact is that free software is also available without paying any money, more or less. So you can get the software at a much lower cost.
More important perhaps is the freedom from control. With proprietary software, you don't really possess the stuff you create with the software, because the company that makes that software keeps insisting that you're pirating their software, and then you have to pay even more money. Or they change the way your documents are stored. If you try to open a document that you stored in a word processor ten years ago using proprietary software, you're probably going to have a hard time opening it again. Whereas with free software, it is always maintained, because there is no one single company controlling it.
For consumers who are organisations, this matters a lot. For individual end users, it doesn't matter that much. But organisations, especially governments, schools and public organisations need to have the independence from particular companies. They'd want to be able to choose what software they have. When you use proprietary software, you don't have that choice anymore, because once you start to use it, it's very difficult to change, because you'll have to make sure that all your habits and practices work well with new software.
So it is a big change to move from proprietary software platforms to free software platforms as a big organisation, but the change has to happen only once. And once you've done that, you really have freedom from this control of licensing companies.
You already named textbooks as another example for the open source model. It is currently used for cultural goods and software only. Do you see possibilities for expanding the model to other goods and services? Where is the limit?
I see two boundaries. One is a sort of "hard" boundary, which has to do with the nature of the goods and services. The other is a "soft" boundary, which changes and has to do with the sustainability of people's incentives. The hard boundary is probably the material aspect. The whole model of freedom and free production depends on the fact that you can copy the whole cooking pot. So it's not going to work for cars. I can never make a car in my garage, or even if I can, it's still going to cost a lot of money and I only make one car, and another car is going to cost more to make than one car. Ten articles, ten copies of the same article cost no more than nine copies. So that's going to be a limit.
That said, you might well have open source car blueprints, where people can share their designs for a car, and you have that already in certain types of sports equipment, where power users make changes to their things – such as in surfboarding, skating, even skiing. You have an open hardware movement that tries to publish blueprints for computers. But there's a boundary where it's no longer open because finally, you have to actually produce the item.
The second boundary is more in terms of incentives and sustainability. In 1999, it seemed that open source was working more for technical areas, where power users and developers were the same people, and less for areas where consumers were very different from the developers. But the incentives have expanded, so now you have OpenOffice, which works just as well as Microsoft Office. It's purely end user related, the developers aren't busy writing big documents or presentations, so there's not very much overlap. But it works because there is a different infrastructure now: There are companies which are paying for this. There are people who develop this software simply because they want to be famous.
So there are lots of different incentives, and I think the incentives differ from subject to subject. For art, it depends what type of art you're talking about. In music, there's an old model which has always existed and is similar to the "give the software away but make money from services": "Give the music away but make money from performances". And in fact, that's still the main way in which people make money from music.
In my journal "First Monday", we published a paper a few years ago that said that in Germany, there are less than 1.500 people who make a living from royalties for their music, who earn more than 15.000 Euro a year through music sales. Everyone else who is a musician has to make a living from either something else, or from performances.
Ten years ago, there was a very well-written paper about the economics of music by Courtney Love, the partner of Kurt Cobain from Nirvana, that said that even they didn't make money out of recordings, they made money from performances and it would make sense to them to give away their recordings for free so that more people would come and listen to them perform.
For other types of things such as painting, I don't know if the same incentives work. For textbooks, on the other hand, these incentives can work, because here it's teachers who are in the best position to create and modify content, and there's really no reason why teachers should not be doing this. Why one ex-teacher should write a textbook out of which a publisher makes a lot of money then, there's no real reason for that today. It's entirely feasible for thousands, tens of thousands of teachers in classrooms to collaborate together, to share their content and make changes that work best for their classrooms.
Can people actually make a living with Open Source? Will the system ever be self-sustainable? To what extent will it always depend on a classic economy or public funding in the background?
Well, people will always need to eat and have money to spend on things. There are two issues there: How much money do they need? And how much of that money do they need to get out of free and open source software?
In software, it's now pretty economically sustainable. It doesn't get much public funding. People do work in day jobs often, but almost half of the free software developers now earn their money from free software work. And this is especially true of the small share that works on free software development most of the time.
The fact is that with these kinds of contributory models, most people don't contribute a large amount. In free software, most people contribute less than two hours a week. You don't need to make much money out of two hours a week. Nor do you have to have very strong reasons to do two hours a week.
If you look at a typical free software project, about 70% of the software is written by a small group of people who contribute 30 to 40 hours a week, and these people typically have very good incentives, whether they're students at university – they don't need to worry about earning money, they're doing this as a learning experience or to build experience for work later because proving that you have done open source projects helps you get a job –, or they're people in companies who are doing business for their companies, or they're independent consultants, or whatever.
But that 70% is not the end product. To be successful, you need the other 90% of people in the community to write the remaining little bits. And they are written by people who are only contributing very small amounts of time.
Interview: Sebastian Deterding