3 ms·

I think lemma 4 glosses over the immense variety of ways computation can be expressed in higher level languages. At one end of the spectrum is source code that

I think lemma 4 glosses over the immense variety of ways computation can be expressed in higher level languages. At one end of the spectrum is source code that’s so poorly written it’s practically impossible to understand or edit. OTOH, well written code with good abstractions can make changes trivial; it can also document relevant information and educate engineers. There’s a lot of value and creativity embodied in well written applications beyond the derived object code and those qualities are (in my mind) distinct from compiled algorithms.

throwawaygh 5y ago
> There’s a lot of value and creativity embodied in well written applications My experience writing patents: 1. Engineer/scientist writes up a very clear and understandable description of the thing. Perhaps including some pseudo-code, diagrams, or derivations along the way, but in any case definitely a high value add description rather than a dump from a git repo. On this we can agree. 2. Engineer/scientist presents this description to lawyer(s), often in the form of a powerpoint. 3. Lawyers then mangle the clean description beyond recognition. There's something very, very, very broken about the patent application/review process. Patents should be written for makers to use and only incidentally comprehensible to legal professionals, not the other way around. After all, isn't encourage dissemination of knowledge the whole point of patents?
- aranchelk 5y ago
  The quoted statement is in support of the idea that software constitutes a creative work and should continue to be be eligible for copyright, not anything in support of patents. I’m not in favor of software patents, but I believe the legal repercussions of successfully convincing people that complex applications should be treated as mathematical algorithms will be almost entirely negative. I imagine the process of creating software patent applications probably varies by org. I agree the review process seems quite flawed, that’s supported by comments I’ve heard from several lawyers as well as high-profile screw ups, e.g failing to research prior art or applying the requirement of non-obviousness both inconsistently and sometimes incomprehensibly. Regarding dissemination of information, yes, my understanding is patents are a trade between the public and the inventor, the inventor receives a period of legally enforceable exclusivity in exchange for disclosing the invention to the public. My opinion is this is overkill in the software industry as we already get exclusivity (and without the need to disclose): we can conceal our source code and assert copyright protections on object code. Conditions are quite different in other industries, for instance manufacture of physical devices.

throwawaygh 5y ago

> There’s a lot of value and creativity embodied in well written applications My experience writing patents: 1. Engineer/scientist writes up a very clear and understandable description of the thing. Perhaps including some pseudo-code, diagrams, or derivations along the way, but in any case definitely a high value add description rather than a dump from a git repo. On this we can agree. 2. Engineer/scientist presents this description to lawyer(s), often in the form of a powerpoint. 3. Lawyers then mangle the clean description beyond recognition. There's something very, very, very broken about the patent application/review process. Patents should be written for makers to use and only incidentally comprehensible to legal professionals, not the other way around. After all, isn't encourage dissemination of knowledge the whole point of patents?

aranchelk 5y ago
The quoted statement is in support of the idea that software constitutes a creative work and should continue to be be eligible for copyright, not anything in support of patents. I’m not in favor of software patents, but I believe the legal repercussions of successfully convincing people that complex applications should be treated as mathematical algorithms will be almost entirely negative. I imagine the process of creating software patent applications probably varies by org. I agree the review process seems quite flawed, that’s supported by comments I’ve heard from several lawyers as well as high-profile screw ups, e.g failing to research prior art or applying the requirement of non-obviousness both inconsistently and sometimes incomprehensibly. Regarding dissemination of information, yes, my understanding is patents are a trade between the public and the inventor, the inventor receives a period of legally enforceable exclusivity in exchange for disclosing the invention to the public. My opinion is this is overkill in the software industry as we already get exclusivity (and without the need to disclose): we can conceal our source code and assert copyright protections on object code. Conditions are quite different in other industries, for instance manufacture of physical devices.