|
|
|
| Authored by: cjk fossman on Wednesday, October 10 2012 @ 03:33 PM EDT |
United States law holds that math is unpatentable.
Since this is a United States legal question, I believe that
holding should apply.
But over and above that, _all_ software is math and therefore
not eligible for patent protection.[ Reply to This | Parent | # ]
|
| |
| Authored by: mbouckaert on Wednesday, October 10 2012 @ 05:10 PM EDT |
Here's the "software is math" debate again.
<asbestos-underwear>
Software is text expressing math. No question.
Running software on a physical processor departs very
slightly from that concept. Very slightly, but it does
depart.
Hardware imposes limitations. The "pseudo-arithmetic" has
been folded back into math's discrete math body. That's not
the point.
The major point is time.
Time is, emphatically, not pure math - pure math knows about
monotonic series, but not about physical time.
Physical time is what makes some algorithms (text which is
math) worth running in a physical processor to solve real-
life problems. Some, even, in what is called real time.
Time imposes constraints over what specific math text is
useful.
Math properties of the various text can be used to select
candidates; but only following the math with a physically-
constrained machine (real time, real limits on the size of
the Turing tape, etc.) qualifies the software text to become
part of a product.
Sure, it's not as simple if that small item called time is
factored in. But I do
not think that this can be avoided. By applying math to a
physical device, "a miracle happens" -- which may or may nor
be patentable.
So yes, I think RSA is patentable. So is quicksort. In
most other cases, the impact of physical time / memory
constraints is so minute as being irrelevant - so these
cases can look at the math text called software, and see
that it is undistinguishable from the real world execution
and is not patentable.
</asbestos-underwear>
---
bck[ Reply to This | Parent | # ]
|
| |
| Authored by: albert on Wednesday, October 10 2012 @ 05:13 PM EDT |
"...RSA is a truly novel and powerful invention that has benefited the
technical world enormously..."
While I agree with the basic tenor of your statement, the difficulty in
quantifying 'novel', 'powerful', and 'beneficial' argues against granting any
math patents. These judgements were made well after the RSA patent, by folks who
are skilled in the art. Not the sort of evaluation one could expect from
judges, juries, or the USPTO for that matter, at the time of the filing.
The first S/W patent is the roach that got into the USPTO, and it became
fruitful and multiplied, and now has infested the whole place. Powerful
remedies will be needed to remove it; more powerful indeed, than the Congress or
the courts can stomach.[ Reply to This | Parent | # ]
|
|
|
|
|
