Re: [Patents] Software is not hardware, social sciences are not natural sciences.

[Jonas Maebe <jonas.maebe@ugent.be>]

On 12 mrt 2004, at 02:43, Russell McOrmond wrote:

> > Software is only mathematics, not physics. That's why the
>
>   Would you suggest that economics is only mathematics?

Economics and software are two completely different beasts. Software is 
merely a way to describe mathematical operations in a way a computer 
can understand them (after all, all a computer does is compute). You 
can of course describe/approximate a lot of things using mathematical 
models, such as things from economics, psychology (e.g. the 
Szondi-test) and natural sciences, but in the end all the computer 
does, is work with numbers. It's merely the user that interprets those 
numbers in one way or another.

Writing software for a computer is similar in principle to writing a 
book for a human: you just describe something, which is interpretable 
for your "target audience". A description of something is not 
patentable subject matter, even if the description is about a technical 
invention (otherwise patent applications would be patentable subject 
matter themselves) or if the description is interpreted using technical 
means (since, as mentioned before, in the end everything is a 
"technical means", including your brain). The technical means 
themselves can be patentable, of course.

Innovations in the way to describe things are not patentable subject 
matter either, otherwise somebody would probably have patented a 
"system and method to circumvent the exclusion of computer programs 
from patentability, by claiming that a computer program interpreted by 
a computer is something entirely different from a computer program as 
such". There's lots of prior art for that now, though :)

Finally, innovations in mathematics are also not patentable subject 
matter.

So the only thing left that's possibly patentable is what you describe, 
but whether or not that's patentable is completely independent from the 
way you describe it.

>   I think there is more to software than just mathematics.  Yes it is
> represented using numbers and symbols which is part of how mathematics 
> are
> defined, but believe there is more to it than that.

If you write a program with instructions for a computer on how to 
operate a drill, you obviously have to know how that drill works. If 
you write a book with instructions for a human on how to operate a 
drill, you also have to know how that drill works. So in that sense 
there is more to programming than just maths, just like there is more 
to writing than literature, spelling and grammar (unless you are 
writing about literature, spelling and grammar).

However, that doesn't change the fact that a program is merely a 
mathematical description of something and that all innovations as far 
as programming are concerned, are innovations in mathematics (and 
logical reasoning, but I consider logic as part of mathematics).

> http://dictionary.reference.com/search?q=mathematics
>
> "The study of the measurement, properties, and relationships of 
> quantities
> and sets, using numbers and symbols."

Frankly, I think this is indeed exactly what programming is all about. 
You have some knowledge, and you have to translate this knowledge into 
"measurement[s], properties, and relationships of quantities and sets, 
using numbers and symbols", because that's all a computer can work 
with.

>   I guess the main point in the context of this conversation is that it
> isn't an application of natural sciences, but of social sciences which 
> we
> don't try to apply statutory monopolies.  Even if software is just
> mathematics, mathematics is not patentable either, and for good reason.

Programming is indeed not an application of natural sciences. I don't 
think it's an application of social sciences either, unless you 
consider mathematics a social science.


Jonas