Monday, November 09, 2009

Court reconsiders software patents

The US Supreme Court just heard oral arguments on the Bilski case:
Chief Justice Roberts told him that he had understood the government’s argument until one of the final footnotes in the merits brief. There, in footnote 30, the Solicitor General’s brief said the Bilski/Warsaw claims might satisfy the Federal Circuit test if they had tied it to “machine implementation,” such as using a computer network.

“That takes away everything you said in 54 pages,” the Chief Justice commented acidly. Stewart backed off a bit, saying that the risk-management notion would not be patentable if a computer were “just used to crunch numbers.” It would have to have something that “gives it functionality.”
What does that mean, that number crunching is not functionality?

The case involves a lower court decision that appeared to abolish business method patents, and computer software patents also. Unless the computer is used for some functionality besides number crunching, whatever that means. The core of the problem is that the courts are unable to separate software from hardware.

I made the point many years ago that the first software patent was not in the modern computer era at all, but the Morse Code patent from 150 years ago. It had a claim for the system of dots and dashes, and the supreme court upheld it. Apparent there was discussion today by the justices about that patent claim.

No comments: