Side effects are included in the absurd legal-patent-weaving language, to get around the pesky thou-shalt-not-patent-abstract-or-natural-ideas rule, but the side effects are not germane to the patent.
There is a logical dilemma with allowing patents based on software side-effects.
Patent attempt #1 is some computation made to effect side effect A.
Patent attempt #2 is the same computation made to effect side effect B.
If you grant both patents, you have acknowledged that the computation is irrelevant to the patent. The side effect, which is often the goal that was set out to be achieved through computation, is not a secret in need of patent protection to encourage public disclosure, so the patent rationale fails.
If you grant only the first patent, you claim that the computation in part affects the novelty of patents, therefore the novelty of math affects the novelty of patents.
This is taken to the extreme where the non-math portion of a patent is a general purpose computer. The meaningful parts of a patent are the math. The general-purpose computer that handles all the side effects is necessary, but uninteresting, and not in the least bit novel.
Most of the time, specific side effects aren't even mentioned in the patents. Here's an MP4-related patent:
The patent, like other software patents, tries to tie the completely functional core (mathematical) idea to a physical computer, but what the trick amounts to is
any arbitrary math function + specification of a general purpose computer + money => patent
(insofar as the math function is novel and not trivial in the opinion of a patent examiner)
There is a logical dilemma with allowing patents based on software side-effects.
Patent attempt #1 is some computation made to effect side effect A.
Patent attempt #2 is the same computation made to effect side effect B.
If you grant both patents, you have acknowledged that the computation is irrelevant to the patent. The side effect, which is often the goal that was set out to be achieved through computation, is not a secret in need of patent protection to encourage public disclosure, so the patent rationale fails.
If you grant only the first patent, you claim that the computation in part affects the novelty of patents, therefore the novelty of math affects the novelty of patents.
This is taken to the extreme where the non-math portion of a patent is a general purpose computer. The meaningful parts of a patent are the math. The general-purpose computer that handles all the side effects is necessary, but uninteresting, and not in the least bit novel.
Most of the time, specific side effects aren't even mentioned in the patents. Here's an MP4-related patent:
http://patft.uspto.gov/netacgi/nph-Parser?Sect2=PTO1&Sec...
The patent, like other software patents, tries to tie the completely functional core (mathematical) idea to a physical computer, but what the trick amounts to is
any arbitrary math function + specification of a general purpose computer + money => patent
(insofar as the math function is novel and not trivial in the opinion of a patent examiner)