4 ms·

I'm betting that in many worlds your "pretty much any programmer may implement" goes for nearly all MechE patents too, civil maybe as well, and plenty of EE pat

I'm betting that in many worlds your "pretty much any programmer may implement" goes for nearly all MechE patents too, civil maybe as well, and plenty of EE patents I'm sure. In reality, nearly all patents are just ::at best:: a year or two of an engineering team crunching on a problem using existing methods - usually far less. It's difficult to say which are which though - so maybe patents are really what we want at all for nearly all things?

Vivtek 14y ago
Yeah, but I don't mind monopoly protection for something somebody worked a year on. Software patents are granted for many times the effective lifetime of a technology, for something you or I could do in a week based on ideas we've all had because they're natural solutions to their problems. Like patenting the use of a hinge for a door instead of patenting how to make a hinge, and the patent being granted for the entire time humanity uses doors at all.
- coopdog 14y ago
  Exactly, you can't patent 'the cure for cancer' without describing the cure, but somehow you can patent 'swipe to unlock'. If it didn't take serious research and give the reader a decent insight that they wouldn't have had just by reading the title of the patent then it shouldn't be patentable. The equivalent of describing the molecule for a drug patent is providing the source code, which is already covered under copyright protection. Sounds like a good case to make software patents redundant
- adavies42 14y ago
  note that both "car" and "computer" were patented at the concept level (both causing huge messes, incidentally). of course, those patents expired long before the relevance of the concepts....
- nickpinkston 14y ago
  I think we really need to consider what patents are for - because "somebody worked a year on" is waaaay to broad. This is about helping society right? Working a year isn't a big investment for a company - so if it's such a great invention they'd easily be able to recoupe their costs. This goes the other way too - Edison's Lab developed the first phonograph supposedly in months - yet that's far more on the "deserving" side of patents. Time isn't the best measure, but we have to consider what an actual good one would be. Giving at patents with relatively little oversight leads to a ton of prevented inventions, but how could we possibly inspect patents to a degree that we'd be able to. I haven't seen a plausible explanation that this search problem is even tractable.
adavies42 14y ago
one of the things someone pointed out to me most of a decade ago is that more and more things are becoming algorithmic--or perhaps we should say, being revealed as being algorithmic. take some classic physical patent like the Bessemer process for steel--given a suitably general collection of industrial robots and reconfigurable equipment, "make steel like this" is "implementable" in software. or any random machinery patent--"be this specific kind of rotary engine" is software if your hardware is a few trillion nanites. what exactly the implications are for the patent system, i'm still not sure....

Vivtek 14y ago

Yeah, but I don't mind monopoly protection for something somebody worked a year on. Software patents are granted for many times the effective lifetime of a technology, for something you or I could do in a week based on ideas we've all had because they're natural solutions to their problems. Like patenting the use of a hinge for a door instead of patenting how to make a hinge, and the patent being granted for the entire time humanity uses doors at all.

coopdog 14y ago
Exactly, you can't patent 'the cure for cancer' without describing the cure, but somehow you can patent 'swipe to unlock'. If it didn't take serious research and give the reader a decent insight that they wouldn't have had just by reading the title of the patent then it shouldn't be patentable. The equivalent of describing the molecule for a drug patent is providing the source code, which is already covered under copyright protection. Sounds like a good case to make software patents redundant
adavies42 14y ago
note that both "car" and "computer" were patented at the concept level (both causing huge messes, incidentally). of course, those patents expired long before the relevance of the concepts....
nickpinkston 14y ago
I think we really need to consider what patents are for - because "somebody worked a year on" is waaaay to broad. This is about helping society right? Working a year isn't a big investment for a company - so if it's such a great invention they'd easily be able to recoupe their costs. This goes the other way too - Edison's Lab developed the first phonograph supposedly in months - yet that's far more on the "deserving" side of patents. Time isn't the best measure, but we have to consider what an actual good one would be. Giving at patents with relatively little oversight leads to a ton of prevented inventions, but how could we possibly inspect patents to a degree that we'd be able to. I haven't seen a plausible explanation that this search problem is even tractable.

adavies42 14y ago

one of the things someone pointed out to me most of a decade ago is that more and more things are becoming algorithmic--or perhaps we should say, being revealed as being algorithmic. take some classic physical patent like the Bessemer process for steel--given a suitably general collection of industrial robots and reconfigurable equipment, "make steel like this" is "implementable" in software. or any random machinery patent--"be this specific kind of rotary engine" is software if your hardware is a few trillion nanites. what exactly the implications are for the patent system, i'm still not sure....