Re:USPTO likely to adopt 'peer-to-patent' (Feb 2010)
by IgorPartola on Tuesday October 12, 2010 @02:26PM
(#1784347)
If you build a physical calculator that can add 2 and 2 in a novel way, you get a patent for the way you've built the machine. By the same token if Intel figures out how to build a quantum CPU they get a patent. You don't get a patent for 2 + 2 = 4, and Intel cannot patent the quantum mechanics formulas, just the blueprints of their device. As for (4), patents serve a purpose in some industries, etc.
1) How would you classify a non-reprogrammable hardware device that performs a computation?
2) Would it matter to your classification if it was programmed through the simple arrangement of wires, rather than some source code?
3) How is an arrangement of levers that performs a computation different? Would you argue that a specific physical machine capable of performing a computation is patentable, while the algorithm itself is not?
3b) Since computers are a form of universal machine, if the answer to (3) is yes, then what constitutes the specific physical machine, where the machine is specified in software?
4) Finally, would you simply argue that nothing should be patentable, in which case splitting hairs is irrelevant?