[Patents] Re: [Patents] Re:[Patents] Prosecution

[Bernard Lang Bernard.Lang@inria.fr]

On Thu, Dec 09, 1999 at 11:50:55PM +0000, Jean-Jacques de Jong wrote:
> "Jesus M. Gonzalez-Barahona" wrote:
> 
> >         As far as I know, for violating a patent you have to provide a
> > *device* which implements the patentes technique. Perhaps an
> > executable program can be conseidered a device. But I don´t see how
> > source code can. And if it ever is, I don´t understand how a frontier
> > can be delimited between fair technical explanation of the technique
> > and providing an actual device.
> > 
> >         Perhaps this could be a line of defense for libre software, in
> > case software patents are accepted everywhere. If distributing source
> > code can be considered just distribution of knowledge, libre software
> > distributors can just distribute programs with patentes techniques as
> > source code. The user who wills can compile the program, and (perhaps)
> > buy a license the patent (although perhaps for private use this is
> > also unneded).
> > 
> 
> 
> This is indeed a very interesting position and it's what I would argue.
> There is no case law, I believe not even in the US, so we must wait and
> see. The problem is that someone must take the risk. Maybe with the
> backup of a company or organization that would like to probe the
> grounds...
> 


Jesus is right on the issue. Is the dissemination of a program a
patent infringement. Currently, it is unclear, though I do suspect
that, at least in the state, you would be prosecuted for distributing
a program with a patented algorithm, for eample RSA or LZW.

And the current attempt in Europe is to patents "programs as such", i.e.
information on a physical medium (floppy, CD, paper ???).
But as Jesus is saying, that is only the description of a process:
  - identical to a mathematical description 
          (programs are mathematical proofs)
  - identical to a precise enough patent (as required by law)

  So writing books or disseminating the patent becomes an infringement.
Absurd ... sure. Where is the line.

  The point is:

  1- a program is information, and information should not be
patentable.

  2- information should not even be contributory infringement, since
the purpose of patents is to disseminate information.

  Hence patenting should only concern programs embedded in a physical
device or process that makes use of them. As has always been the case
with math formaulae.

-- 
Bernard.Lang@inria.fr             ,_  /\o    \o/    Tel  +33 1 3963 5644
http://pauillac.inria.fr/~lang/  ^^^^^^^^^^^^^^^^^  Fax  +33 1 3963 5469
            INRIA / B.P. 105 / 78153 Le Chesnay CEDEX / France
         Je n'exprime que mon opinion - I express only my opinion
                 CAGED BEHIND WINDOWS or FREE WITH LINUX