From endres-sifi@t-online.de  Wed Dec 27 06:19:11 2000
Date: Wed, 27 Dec 2000 07:19:11 +0100
From: A +W Endres endres-sifi@t-online.de
Subject: [Patents] patent pools enforcing per-copy-licensing mode of software
 development

Good morning, Mr. Lang,

Bernard Lang schrieb:

> I do not know Parnas'position.

It is really a side issue. Parnas' most recent publication addressing this subject
is in IEEE Software Nov./Dez.1999, p.19-30. This is mainly about teaching
Software Engineering (SE) and suggest that it should be taught by an engineering
department rather than a science department. It also touches on the question
whether SE is only a part of Computer Science, its common demoninator,
or something different.

>  Inventors never had an exclusive right to their invention, but were
> GRANTED a privilege of monopoly, in return for some contribution to
> society.  You are the one who wants to change things.  The only
> purpose is the public interest... and public interest has to be
> exhibited when extending the range of patentability.

Of course, a patent only gives a limited monopoly for making money
from an invention. I do not want to change that. The patent system
is intended to motivate the inventor to share his knowledge with society.

>    Patents is not about a right to use, but about a right to exclude
> others, even if they discover things independently and can prove it.
> It is about restricting the liberty of others.

Yes, it restricts the liberty to make money from somebody else's
invention. It is assumed that re-inventing a known thing has little value
for society. Therefore publishing of inventions is encouraged. What
am I missing? Should I be more in favor of re-inventions?

>  Hence would you advocate patenting mathematical proof, or keeping them
> secret.  If not, could you explain why ?

I am afraid this is an hypothecial discussion. An invention combines knowledge
from an economically relevant domain in some novel way to solve a problem.
I doubt whether this is the same as a mathematical proof.  If it would be,
I think that the proof (as such) would indeed have the same function in a
patent application that is given to a program. As you know, it is not the expression

(of the solution) that is being patented, but the content (of the solution) as a
recipe for others to use. It has to be understood by the "average" professional
in the field. From this (and the legal tradition) comes the bias for natural
language.
In the case of computer implementable inventions this recipe can be illustrated
by an accepted design notation (where computer scientists would have to agree
upon) or by a program in any popular language (maybe even in Prolog or other
non-procedural notations). This illustration has the purpose to convince
the examiner (representing society) that the recipe works. If a mathematical
proof can serve this purpose, that is OK with me. Again, patenting can avoid
that a solution (and its illustration) is being kept secret.

>    Last point: why would engineers be privileged to own their
> discoveries, and not mathematicians ?

I have nothing against mathemations owning their inventions or engineers
making discoveries. I only said that scientist are more likely to make
discoveries, and engineers are more likely to make inventions. In my
vocabulary, I use discovery for something that pre-existed and was not
created by an individual or group.

> I wish, since you are so eager that everything is taken into
> consideration, that you had been more vocal when only those in favor
> of patenting were the only ones speaking.

I try my best.

Sincerely, Al Endres






