So what did matter about open source? Not “freedom”—at least not in Stallman’s sense of the word. O’Reilly cared for only one type of freedom: the freedom of developers to distribute software on whatever terms they fancied. This was the freedom of the producer, the Randian entrepreneur, who must be left to innovate, undisturbed by laws and ethics. The most important freedom, as O’Reilly put it in a 2001 exchange with Stallman, is that which protects “my choice as a creator to give, or not to give, the fruits of my work to you, as a ‘user’ of that work, and for you, as a user, to accept or reject the terms I place on that gift.”

This stood in stark contrast to Stallman’s plan of curtailing—by appeals to ethics and, one day, perhaps, law—the freedom of developers in order to promote the freedom of users. O’Reilly opposed this agenda: “I completely support the right of Richard [Stallman] or any individual author to make his or her work available under the terms of the GPL; I balk when they say that others who do not do so are doing something wrong.” The right thing to do, according to O’Reilly, was to leave developers alone. “I am willing to accept any argument that says that there are advantages and disadvantages to any particular licensing method. . . . My moral position is that people should be free to find out what works for them,” he wrote in 2001. That “what works” for developers might eventually hurt everyone else—which was essentially Stallman’s argument—did not bother O’Reilly. For all his economistic outlook, he was not one to talk externalities.