Re:Inglourious Software Patents
by kenjackson on Wednesday January 19, 2011 @05:28PM
(#2121524)
I'm not sure what portability has to do with patents? If anything I'd think the opposite to be true. Patents seem more necessary when the barrier to replicate is low. If I have to build a $500M fab that will take 3 years to build, patents seem less necessary, than if I can simply read your code (or disassmble) and copy your logic in a week (or automate the process to copy your code, but do some semantic preserving changes to not violate copyright law).
You don't need patents to protect a specific implementation - code is copyrightable, and stealing your particular code (or even creating a derivative work like your "semantic-preserving changes" comment suggests) is illegal even without patents.
In order to avoid running afoul of copyright law you'd need to do a vacuum reimplementation based just on the description/specification of the algorithm - all that patents achieve is making such vacuum reimplementation also illegal, and it's up in the air whether that's even a good thing or not.
Re:Inglourious Software Patents
by kenjackson on Wednesday January 19, 2011 @06:08PM
(#2121685)
You don't need patents to protect a specific implementation - code is copyrightable, and stealing your particular code (or even creating a derivative work like your "semantic-preserving changes" comment suggests) is illegal even without patents.
I think the case you're interested in is this one: http://en.wikipedia.org/wiki/Computer_Associates_Internation....
Based on trials that have used the filtration technique, if I know the issue ahead of time, I can workaround it.
Reading the actual page (http://en.wikipedia.org/wiki/Computer_Associates_Int._Inc._v...), the "workaround" is a clean-room reimplementation. Unless you were referring to something else?
Re:Inglourious Software Patents
by kenjackson on Wednesday January 19, 2011 @07:44PM
(#2122004)
That's how they did it in this case, but the precedent for determining infringement uses filtration, not the process by which you created the implementation (clean room or otherwise).
Of course, if you do walk this close to the line, you probably don't want to leave a paper trail that you're doing so.
Re:Inglourious Software Patents
by notahacker on Wednesday January 19, 2011 @05:50PM
(#2121613)
If I have to invest $500M and three years to test my hypothesised process improvement, which my competitors can implement with much lower sunk costs if it works, I'm not going to get the funding to do it.
I shouldn't need funding to refactor my code.
Also, software became patentable in 1991 when a lower court overruled US Supreme Court precedent that software for general-purpose computers is not patentable. Some people cite Diamond v. Diehr (1980) as enabling software patents, but that contradicts what the court opinion actually says.