Nice comment/summary from Tim.

On the other hand, the statement “peer production is the most effective way to produce software” is not obviously true. In fact, I’m reasonably certain that with existing economic conditions and methods for organizing peer production, it is false for at least some categories of software. More to the point, there’s no logical connection between these statements. One can perfectly well believe the former while rejecting (or at least being skeptical of) the latter.

The iPhone example simply proves the latter point. So far, top-down, commercial development processes are the best way to produce phone software. But there’s no reason the iPhone should shake your “confidence in freedom.” The iPhone would be an even better product if the software were released under the GPL. I just happens not to be in Apple’s interest to do so.

There’s an obvious parallel to Jefferson’s work. The end is liberty. The means include revolutions, checks and balances, democracy, federalism, etc. The fact that one of the means proves not to work (violent revolution in France, the Articles of Confederation) doesn’t in any way discredit the value of the end. It just meant they had to be more determined and clever in how they pursued it.

By the same token, there is more than one route to free software. Samba, for example, cloned an existing proprietary protocol rather than starting a new one. The Mozilla project persuaded Netscape to subsidize its efforts in a way that probably wasn’t in Netscape’s interests but was very much in the interests of the free software community. The fact that the free software movement hasn’t yet found a similar method for conquering the cell phone market doesn’t demonstrate that there’s anything wrong with freedom. It just demonstrates that freedom is sometimes hard to achieve.