Slashdot Mirror
Alice Is Killing Trolls But Patent Lawyers Will Strike Back
snydeq writes The wheels of justice spin slowly, but they seem finally to be running software patents out of town, writes Simon Phipps in his analysis of how Alice Corp. v CLS Bank is becoming a landmark decision for patent cases in the U.S. 'In case after case, the Court of Appeals is using Alice to resolve patent appeals. In each case so far, the Court of Appeals has found the software patents in question to be invalid. ... As PatentlyO points out, the Alice effect is even reaching to lower courts, saving the Court of Appeals from having to strike down patent findings on appeal.' Although the patent industry broadly speaking sees the Alice verdict as a death knell for many existing patents, some expect Alice to turn software patents into 'draftsmen's art because as you and I have seen over the years, every time there's a court ruling it just means that you have to word the patent claims differently.'
Alice is a big deal. It's already dealt tough blows to some patents in currently pending cases. This is mostly a good thing. And the patents that Alice affects most are sort of the worst of the worst.
I want to address the last point:
This may well be the case. But I don't see that as a particularly bad issue. Among other things, the worst offenders were patents that issued between 1996 and 2006 where there is a huge number of "do such and such old thing on a [computer | web | network| mobile device]" as if the "such and such old thing" was suddenly now a patentable thing by virtue of a new platform. Alice will largely undo those patents.
I also think that the "draftsman's exercise" is likely to bring added meat to claims. It might not sound like much, but every single word that gets added to a claim is one more point of possible differentiation.
If they have to word the claim to actually claim something innovative instead of just a way of implementing a solution to a requirement that anyone can come up with they're in trouble.
My expectation is that Alice Corp will not be used to invalidate all software patents, but it will be the end of most business method patents. Moving forward patentable inventions will again need to be primarily based on technological advances rather than business benefit provided by the solution.
Over the past 160 years, the Supreme Court has decided dozens of cases involving subject matter eligibility. This goes back to a time before Samuel Morse tried to broadly patent the telegraph. (He wanted a patent that covered any and all methods of using an electromagnetic signal to transmit at a distance a signal representing characters). I see the Alice Corp decision as part of that unbroken chain of decisions going back generations. In Alice Corp, the Supreme Court added only incrementally to those cases. The importance of the case, however, comes from the fact that the lower courts (namely, the Federal Circuit) had deviated significantly from that line over the past few decades. I wrote about this here: http://patentlyo.com/patent/20...
If some guy named Bob plays a part in getting Citizens United overturned, you have proof that you are a character if a poorly-written novel.
the preceding comment is my own and in no way reflects the opinion of the Joint Chiefs of Staff
One caveat to the "if morse had won comment" is that our system allows for multiple overlapping patents. Thus, although Morse had the general concept of transmitting characters, another person could still patent a particular implementation that was inventively different that what Morse had accomplished. A third party wanting to use the new implementation might then need to get permission from both patent owners.
I lost my faith in the patent system when I got the text corresponding to my first patent, as it had been redacted by the patent attorney based on my disclosure. It was so vague and wooly that I could not understand what the claim was all about. And I had written the disclosure myself.
Many software patents fail the obviousness test or get by because of ignorant examiners. A disproportionate number of software patents are just doing something that is commonplace in the physical world but only now "with a computer" or "over the internet" as the innovative step. Then comes the standpoint that algorithms aren't patentable while many software patents are only describing an algorithmic process.
The USPTO also has a massive conflict of interest in being dependent on funding from application and maintenance fees. They milk the gravy train of software patents for all it's worth even if it isn't in the public interest to do so as demonstrated by all the NPEs that collect patents for the sole purpose of litigation. It's easiest for the trolls to do this with software patents because it requires little expenditure of effort to come up with something "novel" that will get accepted by an examiner.
I am becoming gerund, destroyer of verbs.
Patents don't disclose source code. So they don't teach a programmer how to "put the bits of plastic together". There is not enough utility in the patents. When creating a finanical application I am better off reading a textbook about finance than looking at patents. The textbook teaches the same thing but uses langauge a programmer can understand. If the patents disclosed souce code then there would be more fans. Patents could become an alternative to open source. The goverment could be using patents to build a library of code that future generations can build applications off of. Instead we get lawyers poorly paraphrasing existing knowledge.
Along with the longer articles mentioned, here's a one sentence summary of the ruling:
Adding the words "on a computer" doesn't change the patentability of a supposed invention.
In Alice, someone basically tried to patent "do escrow on a computer". The court ruled that "do escrow" isn't new or patentable, and adding the words "on a computer" doesn't change anything.
Some in the Slashdot crowd may be tempted to, through wishful thinking, add meaning that the court rejected. The court did NOT rule that having that you can't patent anything that can be done on a computer. They ruled that:
(not patentable) + (on a computer) = (not patentable)
The wording of the opinion also suggests that probably:
(patentable) + (on a computer) = (patentable)
In other words:
X + (on a computer) = X
They said that whether or not it's done on a computer doesn't change the patentability, if the computer part is standard, normal computing processes on a generic computer.
That implies that a new invention which uses a computer in a new, different, and useful way may very well be patentable. So for example it leaves the door open to the idea that a method of doing calculus on the GPU instead of the CPU might have been patentable a few years ago - that was a new, inventive way of using the computer, different from how computers had been used before. Alice talks specifically about "wholly generic computer implementation" as not adding anything to the application.