Slashdot Mirror
The Supreme Court Doesn't Understand Software
An anonymous reader writes We had some good news yesterday when the U.S. Supreme Court invalidated a software patent for failing to turn an idea into an invention. Unfortunately, the justices weren't willing to make any broader statements about the patentability of basic software tools, so the patent fights will continue. Timothy B. Lee at Vox argues that this is because the Supreme Court does not understand software, and says we won't see significant reform until they do.
He says, "If a sequence of conventional mathematical operations isn't patentable, then no software should enjoy patent protection. For example, the 'data compression' patents that Justice Kennedy wants to preserve simply claim formulas for converting information from one digital format to another. If that's not a mathematical algorithm, nothing is. This is the fundamental confusion at the heart of America's software patent jurisprudence: many judges seem to believe that mathematical algorithms shouldn't be patented but that certain kinds of software should be patentable. ... If a patent claims a mathematical formula simple enough for a judge to understand how it works, she is likely to recognize that the patent claims a mathematical formula and invalidate it. But if the formula is too complex for her to understand, then she concludes that it's something more than a mathematical algorithm and uphold it."
He says, "If a sequence of conventional mathematical operations isn't patentable, then no software should enjoy patent protection. For example, the 'data compression' patents that Justice Kennedy wants to preserve simply claim formulas for converting information from one digital format to another. If that's not a mathematical algorithm, nothing is. This is the fundamental confusion at the heart of America's software patent jurisprudence: many judges seem to believe that mathematical algorithms shouldn't be patented but that certain kinds of software should be patentable. ... If a patent claims a mathematical formula simple enough for a judge to understand how it works, she is likely to recognize that the patent claims a mathematical formula and invalidate it. But if the formula is too complex for her to understand, then she concludes that it's something more than a mathematical algorithm and uphold it."
Everything to do with money
Get thee glass eyes, and, like a scurvy politician, seem to see things thou dost not.--King Lear
The question should never have been "why shouldn't they be able to patent it", it should always be "why should they be able to patent it". The fact is that a patent is granted at the expense of the many for the benefit of the few, so there should always be a damn good reason for something to be granted as a patent.
Back in the day patents were put in the Constitution to advance the arts and sciences. Medieval guilds protected knowledge which held humanity/society back, so it was preferable to give limited government protection in exchange for opening up the knowledge (so the next generation can have at it, I guess).
Having patents for their own sake seems counterproductive in this regard, as a lot can be reverse engineered in the meantime.
In reality, everyone is told by legal not to look at previous patents ever, just in case they do infringe, it's not willful infringement.... patent portfolios protect the huge corps and the trolls, with very little in between, and the really lucrative stuff is kept proprietary anyway.
So it leads one to ask, while wasting time writing patents apps, what is the patent scheme good for really and is it beneficial for society?
Any patented process and device can be described wholly in algorithmic terms.
Described, yes. Software patents (and business method patents, etc.) are the ones which can be implemented wholly in algorithmic terms. You can describe a new manufacturing process with math, but you can't actually manufacture anything until you apply that math by rearranging matter and energy in the real world. It's the application that the patent covers, not the description. The device, not the blueprints.
This ruling simply states what should be obvious, that adding "on a computer" to an abstract concept does not magically transform it into a patent-eligible invention. The patent is still really about the abstract concept, despite the "on a computer" gimmick, and thus remains ineligible.
There should never have been any question regarding the patentability of taking someone else's invention (a computer) and using it for the purpose it was designed for (speeding up the evaluation of algorithms), regardless of the specific algorithm in question. That would be like patenting the use of an off-the-shelf pocket calculator to evaluate 2 + 2. If you can't patent the algorithm on its own—and you shouldn't be able to, since it's pure math—then it makes no sense to be able to patent it "on a computer".
"The state is that great fiction by which everyone tries to live at the expense of everyone else." - Bastiat
The purpose of patents is not to reward inventors for being clever. A patent is an extremely powerful monopoly, against which even independent reinvention does not protect and is a privilege that in a free market economy should be granted with extreme reluctance, because of the negative effects monopolies can have on competition and economic freedom.
We have patents because in some fields inventors may be discouraged otherwise because the financial outlay for R&D is too high and it is too easy to duplicate the invention. Patents thus encourage inventors (or the people who bankroll them) by making the financial risk manageable; in exchange, the invention enters the public domain after a set period of time, so society benefits too. I.e., ideally we have a win-win situation where both the inventory and society benefit.
But when R&D does not require expensive labs, materials, or processes, that rationale disappears; instead, patents are likely to become the tools of rent-seeking and regulatory capture and impede progress rather than furthering it. And when independent reinvention is common –as is the case with computer science – society does not benefit from granting inventors such an extremely broad monopoly. The narrower monopoly of copyright is instead more suitable when it comes to protecting the genuine interests of software developers, because the costs associated with software projects are generally caused by sweat of the brow effort (especially when managing a large project), not the underlying novelty.
Patents are not granted for the benefit of a few. They are granted because: The Congress shall have power ... To promote the progress of science and useful arts, by securing for limited times to authors and inventors the exclusive right to their respective writings and discoveries;
Having said that, I agree that the question should be "why should they be able to patent it."
Because a "data compression algorithm" is more than a mathematical equation. Indeed, outside the material scope of a computer it has no existence, except perhaps as a thought problem.
The idea of mechanically separating grain is not patentable but a machine which actually does so is. And that patent will cover any machine which works substantially the same way, which is to say follows the same process or algorithm. Do you follow the difference?
What SCOTUS said yesterday was that merely adding a computer to something already practiced in the public domain does not remove it from the public domain. It is not patentable. Not new. That should come as a big "duh" moment for anyone who thought otherwise. But the invention of something that didn't exist in a non-computer form and for which a computer is an essential component, well that is patentable. And the patent will cover any computer or other device running it.
Moderating "-1, Disagree" is simple censorship. Have the guts to post your opinion.
The judgement reflects current law. The article uses the weasel word "if" to continue to promote the anti-patent lobby's favorite lie:
"If a sequence of conventional mathematical operations isn't patentable,
Okay, if, but the first half is false. The idea that inventions based on math is simply a falsehood detained to confuse those who a) don't know any better and are too busy or two lazy to read the couple of paragraphs that is the actual law.
The law says what isn't patentable is "the laws of nature, including the laws of mathematics".
That laws of physics aren't patentable. Does that mean that any invention based on the laws of physics is unpatentable? Obviously not. An elevator is an application of the laws of physics. You can patent an elevator design. You can't patent gravity. PageRank is an invention that is an application of the laws of mathematics. You can patent PageRank. You can't patent the associative law of addition.
That's the law. Some people want to change the law, and that's fine. Current law is that you can't patent the fundamental natural laws of a science, and can patent an invention which makes use of the science.
The point of patents isn't to reward them for inventing a new compression algorithm. They can do that by selling their compression software and keeping the algorithm secret (if they can keep it secret)
The point is to reward them for telling the world how it works, so others can, eventually, use the same algorithm in their own inventions, or learn about compression and create a better one (which they may or may not patent and then the rest of us benefit from that as well)
Can you be Even More Awesome?!
Certainly those types of patents are bad, and potentially invalid. The scenario you described, where someone attempts to patent result rather than the mechanism is invalid and should be held invalid. Note that this is true whether the mechanism uses gears to perform multiplication or transistors. That problem is completely separate from and unrelated to software or math. It's a bad patent because there's no invention, just a goal or result.
I don't know that anyone has done an analysis to see whether or not more invalid "result" patents have been issued for bit-based "inventions" than molecule-based inventions. Whether the mechanism is made of wood or of bits, I know patents that are too broad have been issued in both.
Of course it's conceivable that some patent examiners could have had trouble distinguishing between the result of a software mechanism (find quality web pages) and a particular mechanism for doing so (PageRank). "High quality" web pages could be determined by any number of mechanisms and PageRank is just one of many different mechanisms that might be invented.