The good might be the enemy of the perfect with software patents

I have a few concerns, relating to unintended consequences, with a proposed loser-pays bill on software patents.

I quite like the idea of loser-pays, and it is the standard in many parts of the world. And I certainly don’t like the current state of affairs with software patents. But the intended narrowness of the bill may make things worse.

Firstly, it attempts to define a software patent as anything that might be implemented on a computer. But think about the ideas that might be “implemented” on a computer. Any process, any design for a device. Such ambiguity tells me that litigation will in fact become more involved, and therefore more expensive, in arguing this point.

Remember, my (shall we say) skepticism about regulation comes from the fact that law is, above all else, gameable. Regulation does not serve as rules of the road, as one hopes. Rather, it is one more input into a complex system.

My second bit of concern about this bill is that it is based on the standard that the “patent did not have a reasonable likelihood of succeeding”. If that’s the case, why was the case heard in the first place? The case succeeds, or it doesn’t.

To retrospectively claim it never had a chance is a bit of revisionist epistemology. And a great bit of ambiguity that begs for appeal.

Now, perhaps I am naive in that these sorts of standards are common in lawmaking. And perhaps I am too demanding that we have a perfect law instead of one that claims to move the ball some distance forward.

I can’t help but think that an increase in ambiguity is a favor to litigants. A better law would a) not attempt to define a software patent and b) apply a clear loser-pays standard to all patent litigation.

Laws are standards. Here’s why we keep making more: