Slashdot Mirror
(Highly Divided) Federal Circuit Opinion Finds Many Software Patents Ineligible
ais523 writes "The Federal Circuit has divided CLS Bank vs. Alice Corp., a case about various sorts of patents, including software patents. Although the judges disagreed, to a lesser or greater extent, on the individual parts of the ruling, more than half decided that the patents in question — algorithms for hedging risk — were ineligible patent matter, and that merely adding an 'on a computer'-like clause to an abstract algorithm does not make it patentable. Further coverage is available at Groklaw, or you can read the opinion itself (PDF)."
It is worth noting that despite the difficulty garnering a majority, this was a case with particularly good reasons to expect patent invalidation. First, it is conceptually similar to the hedging risk patent struck down in Bilski--i.e. in broad brushstrokes, the closest thing to it that the Federal Circuit would have thought about first was struck down. Second, and the real crux of it, is that the computer just wasn't that important. Third, although not addressed by the concurring opinion, it was about as obvious as one can imagine.
There is no line. Algorithms were never meant to be patented. If you're using an algorithm, and you don't want other people to know what you're using, then it's a TRADE SECRET. Funny thing though, is that anyone can decompile your software, and uncover your trade secret. So, what happened is, everyone tried to get their not-so-secret secrets to be covered by a patent.
It's a corruption of the system. Nothing more, nothing less.
"Windows is like the faint smell of piss in a subway: it's there, and there's nothing you can do about it." - Charlie Br
It looks to me like the judges are finally deciding not against software patents in general, but patents of the type "doing X, but on a computer" or "doing X, but on a smartphone". The judgement is basically that you cannot patent something that's already patented, or is a natural law, or is an abstract concept not specific to any particular application, just because you describe it in an unusual way or put it in a new context, i.e. "turning a page, but on a touchscreen instead of an actual page" or in this particular case, "hedging risk, but by a computer algorithm instead of by bankers". Also not allowed would be "the browning of grain-based spongy material through local application of heat", also known as "making toast".
I sometimes ask revealing, often ignorant-seeming questions. Maybe they're harder to answer than you think.
Right — And nowadays, if you *really* want to protect your secret, you don't even have to ship it to your customer: Just offer it as a Web service. No way to decompile that (although it can, of course, be reverse-engineered).
Given an enough-connected world, trade secret is again tenable for algorithms. Not everything can work depending on a Web service, but it pushes quite far the need.
This is, of course, one of the main arguments for protection against copying, but on the other hand, the fashion industry does not have this kind of protection and there is a lot of innovation there. It is argued that this innovation is in large part a result of the lack of copy protection. The high end must have something new to sell after they are copied. Also, this copying frequently involves some degree of change itself.
35 USC 101: Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Hm. Hmmm. There's this talk about "machine", "manufacture", and "matter" that makes me think whether "process" meant actually meant "technological process" (including, say, the chemical ones). When exactly was this law written? Before or after Turing? :-) Given the fact that algorithms are essentially very complex equations, one would expect that mathematics would have been mentioned if that were the case.
Ezekiel 23:20
Algorithms are not "complex equations" any more than a machine is.
Oh, but Alonzo Church says they are! And I'd prefer Alonzo Church's words to your words any day.
Why should, say, the marching cubes algorithm, which transforms bitmap data into polygonal surface data, not be worthy of a patent when the set of instructions for turning bauxite into aluminum is? Because one uses a silicon chip and electricity and the other uses a pressure vessel and electricity?
I think there are both practical differences (investment levels, for starters: it's quite expensive to experiment with chemical processes, drug manufacture, upscale them, AND comply with medical and industrial regulations, whereas algorithm design is hardly more expensive than any other kind of academic or applied mathematical work) and theoretical differences (example by contradiction: if algorithms indeed *are* patent-worthy, why not the rest of mathematics? And if the rest of mathematics is patent-worthy, why hasn't any country done that yet? Case in point: in my country, anything derivable from math is explicitly ruled out as unpatentable).
Ezekiel 23:20