Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

Okay, here's my argument against computer software patent trolls. Since IANAL, attacks from knowledgeable readers are requested although I can't promise a good response because I know next to nothing about patent law although am learning and if my start up works may have to learn.

Dear Patent Court Judge:

Don't worry, I'm not going to throw rotten tomatoes at you now. And, no, I didn't slash the tires on your BMW 7 Series.

But here's what's wrong with likely most software patents (or patents on devices consisting of routine computer hardware but with some new software).

We start with three parts: (1) Real problem to be solved. (2) Some "abstract" ideas for how to solve the problem. (3) Using the abstract ideas, some software to solve the problem. The users/customers use the software.

Okay, now we understand that solving real problems (1) is important but that we can't patent abstract ideas (2).

For more clarity an abstract idea might be just how to manipulate some data in a way a clerk could be taught to do. We can't patent the clerk or their work, right?

Well, for more, the abstract idea might be some applied math or some of the math of physics or engineering. Since that's abstract stuff, we can't patent it, right? Moreover, before computers, mathematicians and scientists commonly did such mathematical manipulations by hand arithmetic, that is, with paper and pencil. No opportunity for patents there, right?

So, on to the software (3): Assume, as is usually the case, the software is just something routine (for software) to have a computer do the data manipulations specified by the abstract idea, what we could teach a clerk, what's in the math, or what the person with the abstract idea 100 years ago likely did with paper and pencil. That is, the person with the abstract idea 100 years ago could tell a clerk how to do the data manipulations and not get a patent but now can tell a computer how to do the same data manipulations but get a patent? Something's fundamentally wrong here.

In particular, assume that without the abstract idea, the clerk would have no idea at all how to do the data manipulations and the computer programmer would have no idea at all how to write the software. So, all that's crucial or original is just the abstract idea and not the routine software. That is, between (2) the abstract idea and (3) the computer software, only (2) is crucial or original and (3) is routine.

So, with this scenario, why the heck grant a patent on the computer software (3) when we can't get a patent on the abstract idea (2)?

But not all software is like that. Instead, some software is tricky stuff. E.g., how the heck to backup a relational database while it is being used and changed? One might argue that just how to do that could be, and really should be, written up as an abstract idea and maybe even some form of math and, thus, not be patentable, but sometimes all there is is the computer software. So, maybe, maybe, I'm not fully sure, such software, or a computer with it, could deserve a patent.

Generally, then, I conclude that most software patents have to be based on bad thinking when we can't patent the crucial, logically prior abstract idea.

Finally, let me be helpful: When you get a case of a software patent, just rule right away that the patent is invalid and take a nice long vacation with the time you would have spent listening to nonsense about that case! Take along some good hiking shoes or a good book!



Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: