Slashdot Mirror
Supreme Court Throws Out Bilski Patent
ciaran_o_riordan writes "The US Supreme Court has finally decided the Bilski case (PDF). We've known that Bilski's patent would get thrown out; that was clear from the open mockery from the judges during last November's hearing. The big question is, since rejecting a particular patent requires providing a general test and explaining why this patent fails that test, how broad will their test be? Will it try to kill the plague of software patents? And is their test designed well enough to stand up to the army of patent lawyers who'll be making a science (and a career) of minimizing and circumventing it? The judges have created a new test, so this will take some reading before any degree of victory can be declared. The important part is pages 5-16 of the PDF, which is the majority opinion. The End Software Patents campaign is already analyzing the decision, and collecting other analyses. Some background is available at Late-comers guide: What is Bilski anyway?"
More analysis of the decision is available at Patently-O.
punted.
Bastards. Bilski is invalid, machine-or-transformation is thrown out, but the court explicitly provides no further guidance either way.
The Software Freedom Law Center has a great response up. From SFLC chairman Eben Moglen: "The confusion and uncertainty behind today's ruling guarantees that the issues involved in Bilski v. Kappos will have to return to the Supreme Court after much money has been wasted and much innovation obstructed."
(I hope they'll be providing a deeper analysis later on; the above came out like ten minutes after the decision, so obviously it's just based on the summary of the decision.)
-Karl Fogel
http://www.red-bean.com/kfogel
The judges have created a new test...
No, the judges said that the "machine or transformation" test that the Federal Circuit used to reject Bilski wasn't the exclusive test, but merely an important tool... And then they left it at that, and didn't specify a new test.
The important parts to take away here are that this decision carefully tiptoed away from software, and did affirm that some business methods are patentable.
They've basically declined to create a new test. The "majority" opinion, in its strongest parts, which leaned more in favor of business-method patents, only got 4 votes (Scalia didn't join those parts). The main concurring opinion (by Stevens), which leaned more strongly against business-method patents, also only got 4 votes. Scalia joined a separate opinion by Breyer which emphasized the common points between the two opinions, but leaned slightly against business-method patents. It agreed with the majority that the "machine or transformation test" wasn't the sole test, but still thought it was the main useful one. That opinion also ended up with a list of examples of ridiculous business-method patents that are definitely not legit, so it doesn't seem Scalia would actually uphold most specific business-method patents that came across his desk.
10 PRINT CHR$(205.5+RND(1)); : GOTO 10
So, can I continue getting patents by adding "on the internet" to common sense operations?
You misunderstand the point of that phrase. It's a technique called claim differentiation and shows up typically in a dependent claim, the purpose being to expand the independent claim from which it depends. For example, say my independent claim is a new forward error connection scheme that involves receiving packets from a network, processing them using my revolutionary new error correction system and appending some sort of advanced hash to them [note: I personally haven't invented it yet, so please don't criticize me by saying "but error correction and hashes are known". This is just for example], and then transmitting the appended packets over the network. Then, in a dependent claim, I say, "The method of claim 1, wherein the network is the internet." Because the two claims, by definition, cover different things, the addition of the dependent claim means that the networks in the independent claim include the internet, but also include other networks - local networks, private networks, metropolitan area networks, wireless networks, wired networks, token ring networks, etc. It's purely useful in a litigation sense so that someone can't say "well, you were thinking of the internet when you said 'network' so when I do it over my corporate intranet, it's different and I'm not infringing."
Is also worth reading. Basically, while this is not a hands down win for opponents of software (or business method) patents, the upholding of the older cases (Flook, Diehr, etc.) could give some guidance on future cases that may help them rule out abstract ideas and algorithms. Villa also talks a bit about how the lower courts may see this and how he thinks they may be handling future patent cases.