Monday, June 18, 2007

LinuxWorld questions peer-to-patent

A post on LinuxWorld severely questions the concept of peer-to-patent:

But this patent thing...this is me saying "What the hell are you doing?

What the hell are you doing?

Helping patent trolls with their QA is like going through bandits' ammunition and throwing out the dud rounds for them before they try to rob you.

The IT industry has a patent troll problem, just like the coach industry in England had a highwayman problem. And what the defense can learn from banditry is that the more reliable the weapon system. the better banditry scales. If you're a coach operator facing a bandit with a matchlock, you're more likely to fight at some point, because the match is likely to go out before the bandit can rob you and get away cleanly. But give a lone bandit a flintlock, and that shifts the balance. Introduce percussion caps and it's even more of a force multiplier for the bandit.

Bruce Perens tells us that up to 95% of software patents are invalid and won't hold up in court against a determined defense. The most likely outcome of a patent attack on a free software project is that the project survives. We're now facing bandits with matchlocks.

The Peer to Patent Project would give the bandits flintlocks. Instead of facing bandits armed with patents likely to be bad, we'll be facing bandits who are confident in their weapons. If you think the problem of mostly-bad software patents is bad today, try peer-reviewed patents that are more likely to go off.


Post a Comment

<< Home