Stallman on unitary patent

As I just observed on my Dictionary of IP Law blawg, Richard Stallman talks a lot of sense, though perhaps the message gets confused because a lot of rather hysterical people follow him. But each day I am less convinced of the benign effect of any part of what we know as the intellectual property system, and I certainly don't believe that software patents should be encouraged, so I lean some way towards his position. And the reason I was posting to the other blawg this morning was because of an absolutely spot-on comment about the utility of the portmanteau expression "intellectual property" appended to a recent piece in the Guardian about the dangers of patent inflation if and when we get CRAP, or the Unitary Patent as it is officially known. I think the unitary patent litigation system also plays an important role in this.

He makes the very good point that granting software patents will be the preserve of the EPO (with national offices also banging out national patents, which will probably become pretty unimportant especially for the sort of undertaking that goes in for gigantic software patent portfolios with which to beat up competitors, whether as first strike or retaliation). National courts will lose the ability to strike them down, but will the unified litigation system be rigorous enough to keep them in check?

A software patent that produces a genuine novel technical effect that itself is a patentable invention doesn't seem to me to be objectionable. What we are talking about here is a machine doing something new, and software controlling it. A patent for such an invention isn't actually a software patent at all, because the software itself is not the subject of the protection. Not an easy line to draw, as we know, but at least we have an idea of the territory in which the border lies.

Mr Stallman makes the interesting point - which, forgive my ignorance, I have not seen expressed so clearly anywhere else - that the EPO searches for such a technical effect in the invention itself, not in the inventive step that makes the invention patentable. The software operating a computer does not have to result in a new machine, just an existing machine doing something new, and there is a world of interesting difference between the two. But I'm not here to explore that, just to draw your attention to an interesting piece on the subject - which I will go away and think about.