Re: [Patents] Software Patents: The Patent Trap

[Bouthier Christophe <Christophe.Bouthier@loria.fr>]

On Fri, Feb 22, 2002 at 11:03:35AM -0500, Gregory Aharonian wrote:
> Yes, you are right.  It is bullshit to view software as anything but
> another field of engineering, and thus a field of engineering that has
> to grow up and learn to deal with the problems and benefits of patents.


Ok, so let's try with other arguments.
Let's try to find true differences between software and other engineering
fields.

First, with software, once you have build something, it costs you 
nothing to give it to others. You even still keep your product.
I don't think that it's the same with, for example, electronics,
chemistry, or others. The fact that software is just information,
so an intangible product, is, it seems, a difference from a 
electronic device (physical product), a medicament (physical product), 
and so on.

Do you agree with that ?

So, when somebody creates, say, a new mechanical device, he won't 
give it to others. In fact, in order to others to get the same product, 
they have to build it from physical resources. But, physical resources
are not free, you have to pay for any physical product. 
For software, as giving it to others is free (as in no money), it seems
normal for people to give what they have done, what they have built.
This was not the case with other engineering products.

Do you agree with that ?

Another point is that you can't have a device that found a new, say,
chemistry formula. You can't have a machine that found a new mechanical
device. But you can have a computer that found a new algorithm for a 
problem. It's called genetic programming. The fact that "invention"
could be found not only by human but also by machine are unique, in the
field of engineering, to software, I think.

Am I wrong ? (I admit that it's not because I don't know other
examples in other engineering fields that they don't exist. I'm
open to any counter example here).


This rises another question : let's imagine that you have a software that
solves some kind of problem with genetic programming. Let's also assume
that the genetic search is reitered each time, so that each time you use
the software, the program try to find a better algorithm, based on the
data it has to proceed.
Finally, let's imagine that some way to solve the problem are patented.
It's possible that, in some running, the program "found" and use a
patented solution. But this solution has never been programmed or
even thinked by the programmer (that's the beauty of genetic
programming). Is the program infriging patent ? 
(For the sake of the discussion, let's assume that the 
solution used left some visible trace in the product builded, so that
everyone can see which solution has been used).

So, is the program globally infriging patent and so unlawful ? 
Even if the programmer has never programmed something that 
infrige the patent ?
This would mean that the very concept of genetic programming is then
unlawfull with software patent, wouldn't it ?

Is the program unlawfull only when it does patented run ?

Is the program lawfull, because it was not progammed to 
infrige patent ?
Then software patents won't have any value : every one that want to 
use a patent would just have to put a little bit of genetic programming
on it to use it.

What is your opinion on this point ?

<I really hope to have a true discussion here between thinking, 
even with different opinions, people>

Thanks for all commentary.

Christophe Bouthier
-- 
By the grace of God Almighty
And the pressures of the marketplace
The human race has civilized itself.
It's a miracle.                      -Roger Waters-Amused to death-