But if he aspires to live in a world of reciprocity, then it makes no sense to limit the possibilities of the exchange to code, excluding money, hardware, or services. Because people value different things, money and markets are the way in which societies achieve reciprocity. The ranters claims of the moral superiority of a gift economy or a barter society are puzzling.I'm somewhat amazed that James DeLong hasn't figured out what Torvalds really meant by now. Tim Lee wrote at Tech Liberation Front about this and covered it fairly well in an area where I would have thought that DeLong would have seen a sufficient explanation to pull a small mea culpa. Guess not. Since I'm in a rehashing mood, let's continue with another rebuttal.
It's a geek thing, and something that DeLong will probably not be able to understand given the difference in perspectives at play here. People like Torvalds, and I include myself and many of my friends in this category, do software development as much for fun as any other motivation. This is why the relationship between open source software and corporate America is so wierd. The software is written because someone decided that they didn't like Product $X and tried to do a better job themselves. Other people jump in and contribute code, art, documentation, etc.
DeLong's definition of reciprocity is different than Torvalds. The latter is more specific than the former. Money does not make a kernel work better, and since Torvald has been hired for a while to work on the kernel by various companies, it makes little sense for him to care about money if he's happy with his job. Torvalds did not start developing his kernel as a business, but as a hobby. If he decides that code is more useful to him, then that's his business. It would seem that IBM, Compaq, Apple, Intel and others that have at some point worked on Linux agree with him that code is a more valuable contribution since he won't license them the right to fork off derivative products from what kernel code he owns.
There is nothing stopping a business from dividing up a pie of a few million dollars between the kernel team's full time members in exchange for the value that they have gotten from it. What DeLong seems to want is the rights to the code in exchange for money, which means ultimately the right to fork the code. Torvalds is not giving that up and I don't blame him.
Leave a comment