|
Authored by: Anonymous on Tuesday, January 01 2013 @ 12:42 AM EST |
"Diamond v Diehr was about a process to transform rubber."
Sure. Including a software component.
"The patent was about transforming rubber. You cannot
transform rubber by thinking about it."
And you can't fly a space shuttle by thinking about it, or cause a bounce effect
on a screen that others can see by thinking about it, either.
"You can replicate every software algorithm in your head."
Including the ones that help to cure rubber. And the ones for the shuttle. And
the ones for bounceback. But you don't get useful results, except perhaps
debugging or understanding program flow, by doing any of those things.
"If you couldn't, then the algorithm would not have been able to have been
developed."
Perhaps you aren't familiar with neural net algorithms? Or genetic algorithms?
A lot of algorithms have been developed which perform useful tasks, which were
not even actually conceived of beforehand by humans. Instead, the machine is
constrained to find a solution in a space. So, as a blanket statement, this
doesn't work. In fact, some neural network algorithms are so poorly understood
that a technique called "sensitivity analysis" is used to attempt to
figure out why they reached particular results.
Yet these algorithms are also just computation, often just run on standard
computers (but sometimes run on specialized hardware to make them faster).
"It's how algorithms _are_ developed. We work them out in
our heads."
Usually, but not always, as I just showed.
"I've been writing software for over three decades. Others here have
similar levels of experience."
Yes. Me, for example. I wrote my first program 38 years ago. But argumentum
ad verecundiam is usually bogus, especially when it is being explicitly
attempted by someone who doesn't even know his audience.
"We are trying to help you understand a simple truth."
And yet, you have severe misconceptions about several things. Interesting.
I understood, well before you tried to "educate" me, that most
software patents were bad. I can even sympathize with the argument that all
software patents are bad. But if you are attempting to distinguish Diamond, you
are going to have to do a much better job, because following it closely shows
that it allows much software which does useful stuff to be patented -- at least
when it is doing that useful stuff. Just running the software in the computer
and not doing the useful stuff is much like running any other kind of simulator,
and almost certainly not patentable. But that distinction really doesn't help
anybody.
[ Reply to This | Parent | # ]
|
|
|
|
|