Archive for April 27th, 2009

Sorry, not right. An answer to Raymond’s post on the GPL

I read with great interest the post by Eric Raymond on the GPL and efficiency, “The Economic Case Against the GPL“; because economic aspects of OSS are my current main research work (well, up to the end of FLOSSMETRICS, then I’ll find something new :-) ) The main argument is nicely summarized in the first lines: “Is open-source development a more efficient system of software production than the closed-source system? I think the answer is probably “yes”, and that it follows the GNU GPL is probably doing us more harm than good.”. Easy, clear, and totally wrong.

The post clearly distinguish between the “ethical” aspects of the GPL with the interaction model that is enforced by the GPL redistribution clause; ESR briefly describes a ideal world where:

“If we live in “Type A” a universe where closed source is more efficient, markets will eventually punish people who take closed source code open. Markets will correspondingly reward people who take open source closed. In this kind of universe, open source is doomed; the GPL will be subverted or routed around by efficiency-seeking investors as surely as water flows downhill.If we live in a “Type B” universe where open source is more efficient, markets will eventually punish people who take open source code closed. Markets will correspondingly reward people who take closed source open. In such a universe closed source is its own punishment; open source will capture ever-larger swathes of industry as investors chase efficiency gains.”

So, Raymond concludes, the GPL is either unnecessary or worse anti-economical. The problem lies in the assumption that the market is static, that the end equilibrium will always be optimal, that imbalances in the market are not relevant (only the end result is), and so on. I will start with the easy ones:

  • the market is NOT static. The fact that one production model is (or is not) more efficient is something that can be modelled easily, but is not really relevant when all agents are able to change their own interaction model at will. Many researchers demonstrated for example that in a simple, two-actor market (one OSS and one proprietary), even in the assumption that OSS is superior in every aspect there are situation where the pre-existing network effect will actually be able to extinguish OSS as soon as there is sufficient pricing discretionality by the proprietary vendor.
  • End equilibrium in real-life markets are not always optimal: the existence of monopolies is the most visible example of this fact (and the fact that there is a company that has been found guilty of multiple abuse of monopoly markets should make this clear).
  • The process is as important as the end result: you can become rich after a life of poverty (and receive all your money your last day of life) or have a generally well-off life, constantly increasing and spending what you obtain. What life do you prefer? So, among all the paths that lead to an OSS (in this case, a FLOSS) world, the one that enforces in a constant way an increase of the FLOSS component is preferable to one that, in an hypothetical way, will lead in the end to market domination.

In general, of all the aspects of OSS that are interesting (and there are many), I find the GPL family of licenses as the brightest examples of law engineering, and I believe that a substantial reason for the successes of OSS are dependent on it. Of course, there are other economical aspects that are relevant, and I agree with the fact that OSS is in general more efficient (as I wrote here, here and here). I disagree with both the premise and the conclusions, however, as I believe that the set of barriers created by the GPL are vital to create a sustainable market here and now, and not in an hypotetical future.