That bit is, I think, is the key to unravel whole patents issue.

The other day I was wondering why it was that, somehow, I considered some patents valid and others invalid. For example, I consider many codecs (DivX, JPEG) to be valid patents, but consider patents on certain algorithms to be invalid. Weird, huh? Because codecs are just a specialized algorithm. I just couldn't put my finger to it.

And then it hit me: all the patents I considered valid were actual products. DivX and JPEG were patented, implemented and licensed. The codecs themselves provided a leap in technology, and without that particular codec the product wouldn't exist at all. The patents I considered invalid are either bells and whistles or lack a real implementation. In other words, the creators of DivX and JPEG were making real money by developing and marketing a new technology, whereas Apple (just using the example du jour, please spare me the faboyisms) is trying to shut down competing products by just pointing at a tiny feature and screaming "COPYCAT! COPYCAT!".

The day Apple shows that it's the rounded corners in the iPad that is making them money, and not the fact that they developed a really thin tablet with great battery life and a decent OS, I might reconsider.

He truly understands the field of law and even more, he understands the ramifications that laws and regulations have and how they will affect us in the future. If there was any hope of changing Patent law, this is where its going to start, through Posner.

Tangent, re: Posner and his place in the law:

The thing to keep an eye on is where his approach is helpful and where it isn't. His hammer is the field of law and economics -- he predicts and explains the ramifications of laws and regulations by casting them in economic terms and applying economic theories to them. That's really useful because economics is all about predicting how people will behave and how much value they'll create in various systems, and those are things the law should worry about.

It breaks down, however, in cases where the debate is about how to assign value to behavior. For example, his 2005 discussion of same-sex marriage vs. civil unions weighs the "outrage cost" of the passionate opponents of same-sex marriage against the interests of people who want to get married, coming down in favor of the first group.[1] That might make sense if you assume that one person's outrage at another person's marriage is something that should be valued, like float(10-pissed-off-Tea-partiers) > float(1-happily-married-couple) -- but the whole point of the debate is whether that outrage deserves a voice in our law.[2] His argument added basically nothing to the debate (other than, in my case, a tickling sense of annoyance) because he encoded his answers to the important issues in hidden assumptions about what was valuable and what wasn't.

This is not to drag the man down at all, who I think is amazing. Just to say that when you see a law and economics argument -- "we should do X, because if you view condition Y in economic terms it should lead to condition Z" -- you have to decide not only whether you agree with the internal reasoning, but also whether you agree with the implicit translation between economic terms and reality. Patents, of course, are the embodiment of an economic theory -- so they're the perfect nail to Posner's hammer, and I'm happy to see him out there swinging.

/Tangent

[1] http://books.google.com/books?id=IRMgP6-EjDMC&pg=PA14... [2] In legal terms, whether it constitutes a rational basis or compelling interest for a discriminatory law. In social terms, whether it's something we want to care about and cater to.

I don't see how this could help, especially with software patents. For every nonsense patent, you could always hire an intern to code it up in an app. Even if it gives you zero business value, you can claim that you're using it internally (or including it in something that you're selling to users). The courts are not well equipped to debunk such claims.

I mean really, could you quickly determine which pieces of infrastructure at google are actually not useful? Which pieces of iOS code?

And for hardware patents, it seems even worse. If I come up with an awesome idea for implementing an adder on SOI, it might take years and millions of dollars of fab time to get that idea into production. Should Intel be able to take it for free just because they already have a fab ready to go?

If your plan to eliminate patent trolls can be easily thwarted by spending a few grand hiring an intern, it is not a really good plan.

Also, as noted elsewhere in the thread, trademarks do behave like this: use 'em or lose 'em. They could be similarly gamed, yet it works in practice.

Small entities can only hope to be bought by or licence their IP to larger entities to see the fruits of their innovation brought to actual consumers.

Small entities create innovation across the board and the judge is missing out on them here, imagining most relevant patents as being ascertained by large organizations.