Software Patents.. hate the game, not the player

[robphelan]

I'm not sure this is even applicable - I work in a specialized software environment where a certain function does not exist. It does, however, commonly exist in other programming languages.

I spent several weeks developing my own version - is that patent-able?

I'm thinking about taking it commercial and that it may be worth 5k per installation.

What do you think?

Personally, i'm against SW patents so I don't want to get into an argument on that subject.

 

[DaveSimmons]

IANAL but:

> It does, however, commonly exist in other programming languages.

Then there is prior art. Adding LINQ to Fortran is not patentable once it's been done for C#.

You might be confusing patents and copyrights. You certainly can have copyright protection for your code.

 

[EagleKeeper]

I believe that concepts are patentable, not implementation

 

[robphelan]

like i said, i'm against sw patents in the first place.

It's cheap enough that companies would just buy it instead of trying to create themselves.. i'd only worry about some offshore company ripping off the code somehow and re-packaging it as theirs.

I guess this is where copyright may come in.

 

[Schmide]

Software is Mathematics?

a good read

http://www.groklaw.net/article.php?story=20111005102810176

 

[beginner99]

i'd only worry about some offshore company ripping off the code somehow and re-packaging it as theirs.

Well a patent is only valid for the country it is filed. And software patents are AFAIK not available in Europe. So yeah, the off-shore company could just steal it and sell it in their country and how do you want to prove they copied it?
BTW patenting is pretty expensive...

I believe that concepts are patentable, not implementation

You can patent an implementation but it would be stupid because then someone could do a different one on the same concept and would not have to pay you license fee. So a patent should always be as general as possible but most importantly, the "thing" you patent must be new. Porting an algorithm from one language to another sure is not new if you do not improve it in a significant way, eg. make a newer, better algorithm.

 

[Markbnj]

If we could give you clear advice about whether it was patentable then people wouldn't be complaining so much about the way software patents are examined. It does not strike me as a patent-worth implementation, but then you might get a dumb examiner on a good day and get it awarded.

 

[Sho'Nuff]

I believe that concepts are patentable, not implementation

Your feeling is inconsistent with the U.S. patent law.

<---- Patent Attorney.

 

[Sho'Nuff]

I'm not sure this is even applicable - I work in a specialized software environment where a certain function does not exist. It does, however, commonly exist in other programming languages.

I spent several weeks developing my own version - is that patent-able?

I'm thinking about taking it commercial and that it may be worth 5k per installation.

What do you think?

Personally, i'm against SW patents so I don't want to get into an argument on that subject.

In general, software per se is not patentable in the U.S. Rather, most so-called software patents claim the hardware implementation of software. E.g.,

"A computer readable storage medium having XYZ instructions stored thereon, wherein when executed by a processor said XYZ instructions cause said processor to . . . . "

That type of claim even has its own name - its call a Beauregard claim.

If your wondering how this claim format defines "patentable subject matter" under 35 U.S.C. 101, U.S. case law has held that the implementation of software on a CPU essentially transforms that CPU into a new machine. Thus, the claim is drawn a machine (patentable subject matter under 101), and not an abstract idea (not patentable subject matter under 101).

As to whether your particular concept is patentable, it is impossible to tell without additional details (don't provide them here) and a search of the prior art. But what you are describing certainly could be massaged into a form that would be considered "patentable subject matter" in the U.S.

 

[Sho'Nuff]

You can patent an implementation but it would be stupid because then someone could do a different one on the same concept and would not have to pay you license fee. So a patent should always be as general as possible but most importantly, the "thing" you patent must be new. Porting an algorithm from one language to another sure is not new if you do not improve it in a significant way, eg. make a newer, better algorithm.

The standard in the U.S. is not simply "new." To be patentable, the claimed invention must be novel (new) AND non-obvious. I would argue that use of an alogirthm designed for one language in another language could be considered new, but that it would likely be considered obvious and thus, not patentable.

As to patenting implementations of software, I think you need to be a little more specific. Hardware implementations of software are patented all the time. Patenting highly speciifc implementations, e.g., embodiments wherein code is excecuted by a processor in the context of a particular OS or operating system, for example, would leave the patent open to design around.

 

[beginner99]

The standard in the U.S. is not simply "new." To be patentable, the claimed invention must be novel (new) AND non-obvious. I would argue that use of an alogirthm designed for one language in another language could be considered new, but that it would likely be considered obvious and thus, not patentable.

true, but he could create that algorithm and would not have to pay you a license fee. only reason to patent in the end is to make money by a) licensing it (like ARM) or b) getting a monopoly on a specify Product.

I'm not a fan of patent law but then without out, in certain industries there would be no incentive for innovation. Especially true for pharmacology companies. Analyzing what is in a pill is trivial compared to creating it.