Software Patents Compared to Hard Patents 134
Arie writes "The Slate discusses the obvious differences between patenting an algorithm and a drug. The article introduces the Fence test, which basically says that if you can physically protect your property, you have a case patenting it. In addition, it claims that the burden on a programmer identifying whether he is infringing on a patent or not involves excessive research burden, essentially to the inherent lack of physical boundaries. Obviously the article starts off with mentioning the patent dispute between RIM and NTL."
Thanks for the dupe (Score:0, Informative)
The Concept of Software Patents Makes Sense (Score:2, Informative)
Many of the points in the article were really stupid.
The analogy about fences protecting pieces of real estate made no sense at all -- putting a fence around your land hasn't enabled anyone to patent their piece of property! Furthermore I could do a pretty good job of keeping people out of my algorithm by building a fence around any implementation (for instance, by burying the algorithm in an antifuse FPGA), but again, this has no bearing on patentability.
The fact that algorithms are compositional and therefore harder to draw boundaries around isn't unique to software either. Many other disciplines, including material science, chemical process design, and the article's own pet area of molecular design itself make use of elements that are compositional. Algorithms are compositions of operations on 1s and 0s, just like chemistry is about C, H, O, N, and friends. Or an algorithm is just a Turing machine, and chemistry is about protons, neutrons, and especially electrons. Pick your level of decomposition, but algorithms and physics are still both compositional in a very fundamental sense.
In a more fundamental sense, Computer Science is a branch of Mathematics - number theory in particular. A mathematical fact is not patentable, but the application of a mathematical fact to solve a particular problem is. So you can't patent 1+1=2, but you could patent the application of 1+1=2 to solve a given problem, as long as the application of 1+1=2 to the problem at hand met the standard for novelty. Thus the use of an algorithm to solve a problem should be patentable as well.
This places mathematics on the same ground as the laws of physics. You can't patent gravity, but you can patent the use of gravity for solving a problem.
Patentability 101 (Score:3, Informative)
Patent laws usually require that, in order for an invention to be patentable,
* it must be of patentable subject matter, ie a kind of subject-matter that is eligible for patent protection,
* be novel,
* be non-obvious (in United States patent law) or involve an inventive step (in European patent law);
* be useful (in U.S. patent law) or be susceptible of industrial application (in European patent law).
[* see below for formal conditions such as 'sufficiency of disclosure')]
Usually the term "patentability" only refers to "substantive" conditions, and does not refer to formal conditions such as the "sufficiency of disclosure", the "unity of invention" or the "best mode requirement".
Under United States patent law, inventorship is also regarded as a patentability criterion.
--source: http://en.wikipedia.org/wiki/Patentability [wikipedia.org]
Re:Investment, risk, compensation (Score:3, Informative)
That's completely wrong. The subject of any patent, and the core protected aspect, is an inventive and novel concept. A physical object may incorporate that concept, but even then it is not the invention - it is the "embodiment" of the invention. It is a manifestiation of the inventive concept that the patent actually addresses.
This is evident in the fact that every patent has a difficult-to-ascertain quality called "breadth" - i.e., how broadly does the inventive concept stretch? A patent might claim the concept so narrowly that it only covers a single implementation - this is what we call a "picture claim," where you've essentially taken a snapshot of one embodiment, and gained protection of only exactly that object (and exact replicas.) At the other end of the spectrum, a "frontier patent" may claim a novel concept with an incredible range of embodiments - any implementation that incorporates that concept is covered.
Note: Even aside from software patents, "embodiment" is not limited to physical objects. The patent system has long protected "processes," i.e., sets of actions with novel results. The novelty here is completely in the abstract, functional steps of the process - it does not have to involve any particular object or composition. Of course, this fact is exactly why the argument against software patents is untenable: what is protected is not the actual software - not the code, compiled binary (or script), or the process in execution - but the abstract method that the software embodies... and such inventions have been patentable for over a hundred years.
- David Stein
Re:Investment, risk, compensation (Score:3, Informative)
That's completely wrong. The question of infringement boils down to a comparison of the claims - of the bare, essential, conceptual elements of the patented invention - and the embodiment accused of infringing.
The Court of Appeals for the Federal Circuit has had to correct this misconception dozens of times. Many accused infringers want to point to some feature of the patentee's product in order to make some argument about the proper interpretation (construction) of the claim language. This evidence is irrelevant. The patentee's products are irrelevant. All that matters is the text of the patent.
Now, why is this? It's because the patent covers an inventive concept - the feature that makes the invention novel and useful. Now, it doesn't broadly cover any "idea"; in the words of the USPTO, the idea must be "useful, tangible*, and concrete," rather than an "abstract idea" or a "scientific principle" without a specific use. But it nevertheless covers the class of "ideas" that constitute inventions.
(* Neither "tangible" nor "concrete" here means "physical." "Tangible" means that the invention has some interaction with the real world - e.g., the reallocation of money. But this money need not be physical; it can be virtual, i.e., data - what matters is that the shuffling of this data conveys a result with real-world consequences, i.e., money changes hands. And "concrete" must means "repeatable.")
- David Stein
Re:Investment, risk, compensation (Score:3, Informative)
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."
In case you missed it, "process" is the first one.
This has been the wording of the statute since 1952. And process patents long predate 1952: in this change, the patent law only affirmed the long-standing USPTO and court practice of allowing patents on processes.
In other words, your arguments that patents should only adhere to "things" are over 54 years too late.
- David Stein
Re:Investment, risk, compensation (Score:3, Informative)
What an odd concept. Virtually all chemical research is described by "speech/print." You don't document a chemical engineering technique with pictograms or multimedia; you show it by describing, in words and letters, the steps.
In fact, your distinction would relegate every single patented invention exclusively to the land of copyright. You see, since 1952, patentees have been required to describe their invention in claim language - a single sentence of plain words that summarizes the invention. Even the most complex machines that are best explained through a drawing must still be claimed in plain words. Claims like "I claim the invention shown in Figure 1" are per se invalid, and are rejected out of hand by the patent office. You have to say, "I claim: A machine comprising: a sprocket, a flange connected to the sprocket, a winch connected to the flange,..." By your logic, every one of these claims is a summary of the invention in "print," and so every one of these inventions is only protectible by copyright. Do you see now why your test is nonsensical?
At its extreme I have seen plans for a piece of furniture that has no patent protection on it try to apply EULA copyright restrictions to your ablility to impliment the plans and claim a license fee for every instance of the furniture you build.
Oh, dear. I guess it's time for the biweekly Slashdot Intellectual Property 101 lecture once again...
Many objects, including your hypothetical piece of furniture, have both patentable and copyrightable aspects. Copyright protects original expression; patents protect functional novelty. A chair may have both a distinctive artistic quality that constitutes "artistic expression," and also a patentable support structure. In fact, the same physical element can easily serve both purposes, and so be covered by both kinds of intellectual property law.
In the case of your chair, the "implementation" of the plans is called derivatization. You are taking the artistic work described in the blueprint, and you are creating a derivative - i.e., a physical chair embodying the same artistic expression. It's exactly the same as making a movie version of someone else's novel - what you're doing is "adaptation," which is the novelist's exclusive copyright over his novel.
- David Stein
Re:Investment, risk, compensation (Score:2, Informative)
The original purpose of a patent is to encourage people to invent things which better society as a whole, by offering them a temporary monopoly over their invention in return for its eventual release to the Public Domain for the benefit of everyone. If an invention is any good, it will recoup its development costs and maybe make a little profit but that is a side-effect of the means, and not part of the end.
All means to the same end are equally valid. So maybe we need to look seriously at alternatives to the patent system: we still want to encourage people to invent things which better society as a whole, but is a temporary monopoly really the best way to achieve this?