Solving the Patent Problem
To my mind, patents in software add no value, and really only serve to subject software development to the less than scrupulous (lawyers, that is).
It is starting to look to me as if the OSS world won't ever meet its real potential unless software patents go away. Since the non-OSS world has little incentive to change, it would seem to be up to OSS boosters to drive it.
Much to agree with in there, but in the current political climate both in the US and in WIPO I feel there's little hope of traction in undoing patent legislation, even with heros like Lawrence Lessig fighting the case.
What became apparent (to me at least) at F/OSS this week was the need for research into how to create mechanisms that protect open source projects against patents. In his evening keynote Richard Stallman suggested building an anti-patent 'poison pill' into the GPL such that any attempt to assert patent rights against a GPL-protected project would result in automatic loss of rights to use any GPL-protected software anywhere. One delegate noted, however, that such a provision would effectively prevent a company like IBM, with a history of patent suits, from using GPL software. Ironically, a company like Microsoft with no history to date of patent suits would be able to use GPL software without any problems!
As I suggested before, I think it's necessary to understand what is the real problem patents pose to open source and then address that problem. It seems to me that, to expand Benkler's definition, open source is commons-based, rapidly-iterative peer production. Patents delay or erode the commons, and disincent peers from involvement in rapid iteration. Fighting patent legislation is one way to address the commons, but a seperate mechanism such as use of foundations to own code as a haven against prosecution seems vital to address the iteration issue
I'm not a lawyer, so have no training to construct the defences against patents. But it seems to me this is a vital research area, and I encourage the F/OSS academics to engage in it without delay.
[Also posted at Webmink]