Poll: Patents and innovation

No readers like this yet.
Software patents


After you cast your vote, tell us what you think about the FTC weighing in on patent reform or if software patents are too abstract to patent. Comments welcome.

User profile image.
Rob Tiller is vice president and assistant general counsel for Red Hat, where he manages patent, trademark, and copyright matters. He is a frequent speaker and writer on open source legal issues. Before coming to Red Hat, he was a partner with the law firm of Helms, Mulliss & Wicker, PLLC, where he specialized in commercial and IP litigation.


Far more money is being spent on patent related issues than research. That alone is stifling innovation.

Software is mathematics. Mathematics is not patentable. The whole current process is irrational because the lawyers and judges and legislators WANT software to be patentable and so they keep trying to argue that some mathematics is mathematics and other mathematics is not mathematics. It is inherently contradictory and the current disaster is what we have.

In general, patents, even on non-mathematics are destructive under the current system because 1) anyone accused is assumed guilty and 2) patent defense is so expensive that anyone accused is forced to give up unless they are a huge. wealthy corporation. So, patents can never help the "little guy."

Contributor PolR has written an article on Groklaw which addresses Renee's first point in more depth:

The argument breaks down into steps (1) All software is the implementation of algorithms (2) all algorithms are mathematical algorithms (3) all mathematical algorithms are mathematics (4) the US Supreme Court has said on a number of occasions that mathematics does not constitute patentable subject matter.

Yet, if you examine the body of issued patents, you will find patents where the entire content of the claims is an abstract description of an explicitly "mathematical" algorithm. An example is the Fast Fourier Transform.

It would be better if the US Supreme Court made definite pronouncements on the whole chain of argument. Is (1) true in US law? Is (2) true? And so on, with justifications, so that we can understand their reasoning, and refute it, if necessary. This is where articles such as that by PolR help.

Patenting software is a sensible as patenting the common house brick. More than anywhere else (that I am aware of) software is developed incrementally, idea upon idea. Anything that interferes with that process puts a brake on progress and stifles new ideas before they are even born.

If you state the objective, any student of Computer Science 101 can write the software to accomplish the task. He/she just breaks the problem down into smaller steps and connects the statements in programming language to implement them. It's a one-to-one process and involves no innovation. It is merely translation, a creative work with copyright, but not patentable. Patents are supposed to inspire innovation by revealing the steps in the process. The software-patents we have seen do not do that but merely claim the outcome is an invention. So, software-patents do not accomplish the goals of the constitution and in fact do the opposite, preserve market share against real competition. Further software-patents provide reward for millioins of copies of some software which took minutes of effort to code the particular claimed invention, a gross surplus compared to the guy in the garage building machinery by hand. A patent should not reward an inventor thousands of times more just because he implements his supposed invention on a computer. That makes no sense.

>> any student of Computer Science 101

One major problem with swpats is that anyone can participate. We don't need to find $1 billion capital. We don't have to have a few people waiting for money from the sky. We can have millions of people collaborating and distributing right away and moving on to the next stage of invention. So there is not even the appearance of justification to award the monopoly on the ideas. When you award them, you are denying many smart people from across the spectrum (including many students of all sorts) from recreating the same invention and from adding on much new innovation, though, just the fact these would-be contributions would be made independently of the patent means you are abridging First Amendment original free expression, which further means on this count alone you stand a fair chance to be stifling progress as well. [These are two Constitutional no-no's.]

As if this wasn't bad enough, patents offer an alternative cost to society to trade secret. The real algorithms can be kept secret while the broader usage concepts and techniques can be patented, giving society the worse of both worlds. Where the invention was not that advanced and would or had been discovered by others privately or even publicly (since the USPTO does not read public source code or go through blogs, they award patents even when the actual prior art was already significant), we now have to wait 20 years, during which time others are bound and we still don't know the potentially really clever trade secrets.

But the real kicker perhaps is that the inventiveness bar is so incredibly low: non-obvious to a PHOSITA. This guarantees the progress is not promoted, as the majority of above average folks (from a very large population of participants, sw being practiced for so cheaply) will find themselves in violation of what to them are obvious inventions.

We can resolve all of these issues one by one by adding a number of patches to patent law, or, easier yet, we can just recognize that monopolizing ideas and information (stuff that you can write down, speak to others or send them electronically, and "implement" at a very very low cost) was never the intention of patent law, partly because it is a surefire way to stifle society and abridge the freedoms of many. Congress has never been given authority to dole out patents that stifle the progress, especially when that results in the abridgment of so much communication and expression.

Rob, I don't know how else to get in touch with you, and I think you might want to know about this. I have reduced the BTC patent in the recent Google case to a set of mathematical formulae in the Lambda Calculus. I believe that this would be of use to you in Red Hat's attempt to get the patent declared invalid.

You can find details in my blog at http://paulspontifications.blogspot.com/2011/04/patent-5893120-reduced-to-mathematical.html, and the actual formulae are at http://hpaste.org/46182/patent__5893120

Email me if you want more information.

Thanks for the information. I believe most of these techniques are also documented in computer science texts such as Knuth's <a href="http://en.wikipedia.org/wiki/The_Art_of_Computer_Programming">The Art of Computer Programming</a> which dates from the 1960s. Certainly they are implementable in many programming languages such as C, Pascal and certainly assembler.

I remember teaching high school computer science about 2003 and using linked lists and hashes to make old computers rockets to manage information. The students were absolutely amazed that they could write a few pages of text that would enable a computer to decipher text that had permuted characters within words much faster than they could read. They began with an objective that seemed impossible, but with the addition to their toolboxes of a few easily understood techniques, the whole problem became trivial. The key element was a dictionary revised so keys were made of lists of characters sorted aphabetically and hashes were made of the scrambled words by sorting alphabetically and suddenly the problem was reduced to a rapid search in a tree-structure. Students were amazed that a well-defined problem that would take humans hours to do could be done in a second using a computer. The point it that no creativity was required at all given a few well-know algorithms.

The poll might be interpreted to be asking if some innovation is encouraged in some people when we allow for patents. Many people might answer yes to that question. The important question is, do software patents result in greater or lower overall innovation/inventions v. not having them? In this case, if it helped a few be more innovative but destroyed lots of potential innovation from millions, then society and progress in the art would be worse off.

A patent infringement claim might make some sense if the implementation targeted was not as software fed to a general purpose executing environment accessible to many. In other words, target the patent at most at particular hardware that essentially can only solve that specific problem for the life of the product.

[I don't think even that patent protection would be good if it can cover very few algorithms, at least not when the monopoly lasts 20 long years, but such claims are much more reasonable in contrast to today's swpat claims. We'd much less so be preventing new inventions and independent creation in allowing for a monopoly on a very precise expression of circuit that is generally out of the reach of the vast majority of people. However, isn't this more specific case already covered by copyright law?]

People collaborate with software and on many other types of information products because there is much to be gained. There are many hobbyists and many ways to make money. We also frequently have copyright already in play. And this ability and human need to create something fresh (especially when further enabled with collaboration and sophisticated processing and information access) is more possible in today's Internet world than it ever was.

Patents cannot be a tool for promoting collaboration or even individual creation in itself because they specifically end many types of collaborations, including making people wary of sharing since any particular observer might seek a patent related to what you are sharing, and they disallow many individuals to try and succeed at solving a particular problem that overlaps an existing patent.

Patents are about putting forth a very significant restriction only against large competing capital in environments where that capital is required and otherwise might not be forthcoming thereby denying the natural evolution of progress.

Thus, patents enforced on software development, math, music, creative writing, law, and any of many other "free" information transfer environments, where people necessarily share and collaborate given few barriers, results in stifling and in a very large number of people having various freedoms denied.

Creative Commons LicenseThis work is licensed under a Creative Commons Attribution-Share Alike 3.0 Unported License.