While I generally find myself on the side opposing patents, this article was an interesting read from the pro-patent perspective and I feel made good arguments.
I personally like to know and understand both sides because, in my experience, in 99% of cases no side is 100% right and no side is 100% wrong.
OK, Mr. Goetz is correct for the most part on all counts. And I don't understand enough about the software patents arguments to really comment one way or the other. But what I do know is that the bottom line, and what ultimately turns me against software patents (and drug patents as well) is that this is really all about money.
How does humanity as a whole benefit from a patent holder demanding compensation for the use of their "software invention" when it's duplicated, unknowingly, by an independent hacker working in his basement on a pet project following the rapid uptake of their project by the greater software community? What if the hacker gave the software away for free? How does the original patent holder get compensated?
Or what happens when implementations of useful, but somewhat obvious technical solutions to problems are patented? Like linked-lists, for example? Or a compression algorithm? etc. etc.?
IMHO, just because someone implements something in software first, doesn't necessarily give them the right to own that implementation for some unreasonable amount of time.
I think the problem you're getting at (and I agree with you) is that obvious things are patented all the damn time, and it seems to be especially a problem in software. We need better trained examiners, and it ought to be a high bar.
I don't have any issues with it being "all about money." I do have issues with someone getting a patent on a method that would occur to anyone trying to solve a similar problem.
The legal problem is that most software patents are plans for instructions for how to construct something tangible (albeit that is an abstract view). Which basically means that they lie in some grey area in between written works and designs for something.
I would argue that maybe a good rule is to force patents to applicable only on systems of certain size or complexity. That way algorithms and data structures would not be patentable but a method for mining large datasets would be. Similar to how a new dimension of a nut and bolt is not patent-able but putting a bunch of them together in a innovative way is.
This is basically a semantic argument for how to classify software. It does a fair job of showing that the difference between software and hardware isn't necessarily distinct, quantitative instead of qualitative.
But that's ultimately an academic argument. However you classify software, however you elide the distinction, however you define the terms, in practice the patent system is supposed to encourage innovation, and this argument does nothing, in fact doesn't even attempt to address the growing charge that it does precisely the opposite.
It's not just semantics though. If you're against software patents but ok with patents in general, there needs to be a principled way to draw the line between them. The point of this article is that there really isn't.
[+] [-] jstedfast|14 years ago|reply
I personally like to know and understand both sides because, in my experience, in 99% of cases no side is 100% right and no side is 100% wrong.
[+] [-] wmat|14 years ago|reply
How does humanity as a whole benefit from a patent holder demanding compensation for the use of their "software invention" when it's duplicated, unknowingly, by an independent hacker working in his basement on a pet project following the rapid uptake of their project by the greater software community? What if the hacker gave the software away for free? How does the original patent holder get compensated?
Or what happens when implementations of useful, but somewhat obvious technical solutions to problems are patented? Like linked-lists, for example? Or a compression algorithm? etc. etc.?
IMHO, just because someone implements something in software first, doesn't necessarily give them the right to own that implementation for some unreasonable amount of time.
There must be a better way?
[+] [-] monochromatic|14 years ago|reply
I don't have any issues with it being "all about money." I do have issues with someone getting a patent on a method that would occur to anyone trying to solve a similar problem.
[+] [-] crxpandion|14 years ago|reply
I would argue that maybe a good rule is to force patents to applicable only on systems of certain size or complexity. That way algorithms and data structures would not be patentable but a method for mining large datasets would be. Similar to how a new dimension of a nut and bolt is not patent-able but putting a bunch of them together in a innovative way is.
[+] [-] joebadmo|14 years ago|reply
But that's ultimately an academic argument. However you classify software, however you elide the distinction, however you define the terms, in practice the patent system is supposed to encourage innovation, and this argument does nothing, in fact doesn't even attempt to address the growing charge that it does precisely the opposite.
I ask the more relevant questions here: http://news.ycombinator.com/item?id=2875289
[+] [-] monochromatic|14 years ago|reply