Mike Masnick points to a post detailing a possible model for better handling software patents. If reform that eliminates them is simply not possible, which seems to be increasingly the case here in the US, then this is an interesting idea. In short, it hinges on whether a developer, given a patent, can describe a similar way of solving the problem.
Masnick is curious whether such an idea could be pursued as an experiment. It doesn’t seem too far different from the peer-to-patent experiment though narrower in focus. He’s concerned about whether enough developers could participate to make a difference, given that most have day jobs. The implication is that if they undertake patent examination full time, they’ll become disconnected from the skills and experience that make them valuable.
I wonder if the experiment could be crafted a bit differently, to rely on conscripted developers for a limited time, rather than volunteers working at near full time. Granted, a recruit isn’t likely to have the same motivation as a volunteer but some way of better distributing the cognitive load would seem desirable.
The more I think about it, though, the more concerned I am about gaming. Would the patent office restrict itself to drawing an experts from small software shops or exclusively FLOSS projects? As much as I fall on that end, often against the interests of the biggest software vendors, I think any such experiment has to have some checks in it to deflect potential gaming towards either end of the spectrum of interest.