4 ms·
Software differs from other fields in two important regards: 1. Software is math. Not "does" math, but "is" math. See Google and HNsearch for references to t
by nitrogen 12y ago
Software differs from other fields in two important regards:
1. Software is math. Not "does" math, but "is" math. See Google and HNsearch for references to the endless discussions of this fact.
2. A patent on a device covers an implementation. Patents on software have been used to block entire ideas, regardless of implementation.
Based on these factors, plus many more that have been argued to death, software should be ineligible for patents, and software on a computer of any kind should not be capable of infringing any patent. Failing that, independent invention needs to be a 100% defense against infringement, and evidence of obviousness, especially with software, to release the silent stranglehold that the threat of litigation has on software innovation.
- spacemanmatt 12y agoThe Supreme Court essentially agrees with this in Alice vs. CLS Bank
- rhino369 12y agoAlice hasn't destroyed all software patents, only the egregious ones.
- spacemanmatt 12y agoTo be sure, the ruling only destroyed one patent. It did however create a precedent that can be used to defend from (or more ideally, will prevent) litigation over abstract subject matter.
- bougiefever 12y agoWhat did Bob say? Did he get Alice's message?
- yellowapple 12y agoI bet Mallory would know.
- tbrownaw 12y agoSoftware is math. Not "does" math, but "is" math. See Google and HNsearch for references to the endless discussions of this fact. The laws of physics for macroscopic scales are very well understood. You car is as much "math" as any reasonably complex end-user software package is.
- nitrogen 12y agoIf you want to go down that route, then I would agree with the final conclusion that nothing should be patentable that can be described in any way using known physics and chemistry. But I think it will be necessary to draw a distinction between mathematical descriptions of physical processes (which don't make the processes themselves mathematical), and mathematical implementations, namely software. In other words, math describes physics, but math is the same as software.
- Animats 12y ago"Failing that, independent invention needs to be a 100% defense against infringement" Won't work. That's a state-of-mind issue. Infringers will claim "independent infringement", when in fact they got the idea from the patented work.
- ahomescu1 12y agoSure, and then it would be up to the patent holder to prove intentional infringement. Otherwise, they don't get anything. That's how the legal system works in many other situations, the accuser has to prove their accusations, which is perfectly reasonable to me.
- Animats 12y agoPatent infringement doesn't require intent. An "independent invention" defense rewards infringers who keep their technology secret. If the technology is publicly disclosed by an independent party prior to the patent application date, it's not patentable anyway. But secret independent invention is not prior art. Part of the idea behind patents is to reward public disclosure. After the 20 years are up, anybody can do it.
- ahomescu1 12y ago> Patent infringement doesn't require intent. I know, but that's the whole point: maybe it should (in a lawsuit at least).
- nitrogen 12y agoGit histories and lab notebooks can demonstrate independent invention by showing the process leading to independent discovery.