Re: [Patents] contrefaço n involontaire

[Xavi Drudis Ferran <xdrudis@tinet.org>]

El Sat, Mar 15, 2003 at 02:31:55PM +0100, Jean-Bernard CONDAT deia:
> AU SECOURS. Cessez de me dire qu'il y a inventivité dans les brevets
> logiciels. Ma mère le saurait. Et pçour l'instant, elle a des problèmes de
> livraisons avec DELBART de ses fraisiers... ok?
>

Excuse me if my poor French betrays me, but if I undestood right, you're 
right. No software patent shows inventivity these days, and none will 
ever do. Software creations are problem solutions inside a predefined 
formal model. That can never be inventive, it's all already in the model.
They can be new, they can be brilliant, they can be elegant, but they'll
never be an invention, so they'll never be inventive. 

In traditional fields it is not enough to deduce your solution from a 
model of the world. You must test it really works in practice. This alone
involves a minimal amount of work that in a way should prevent absolutely 
brain dead patents to issue. So in those empirical fields you can possibly
make the patent system work (only a theoretical possibility, not much 
evidence in our current patent systems, but at least some hope). 

By trying to embrace fields in which there is no way at all to find
inventivity (or any economic benefit from the patent system) the
patent inflationists are just digging their own graves. If they keep
insisting in blurring any limit to patentability and just pretending
they need to improve examination and prior art searches, they may
reach a point when the evident obsolence of the patent system will no
longer be solved by constraining its scope to the fields where it can
more or less serve public interest, since the system itself will have
erased any usable limits, and then the only solution will be banning 
the whole patent system, possibly a suboptimal solution, but better 
than the current system or where the patent inflationists are heading.

The reason software patents are trivial is because ideas are trivial 
in logical fields, implementations are hard. In order to have nontrivial 
patents you would have to patent such a specific implementation that 
you would only get an expensive copyright. 

-- 
Xavi Drudis Ferran
xdrudis@tinet.org