RE: [Patents] Linus: kernel development no longer legally possible

["Steve Wilcox" <spwilcox@acm.org>]

So the problem is not software patents per se, but rather the types of
patents are granted.  Broad claims are made and acceded to.

I am neither a lawyer nor a mechanical engineer, but I think that if I
submitted an overbroad patent application for a general widget, they would
laugh at me.  In software, however, they wouldn't.  There must be a reason
for this.

-----Original Message-----
From: PILCH Hartmut [mailto:phm@a2e.de]
Sent: Friday, September 06, 2002 10:50 AM
To: Steve Wilcox
Cc: news@ffii.org; patents@aful.org
Subject: RE: [Patents] Linus: kernel development no longer legally
possible


> If the principal objections to SW patents are (1) obviousness and (2)
> novelty, both according to the standards of the art in question
> (computer science, operations research, statistics, finance and
> accounting and the other information specialties), then why all this
> ruckus about whether software is patentable per se?

The principal objection is broadness, not novelty or non-obviousness.

If swpat were reasonably narrow, nobody would want them, because in order
to get them you have a lot of overhead (such as disclosure, writing an
application, paying fees etc).  Swpat can only be unreasonably broad or
not be.  But broadness violates basic constitutional rights of freedom of
expression.  Even if there was an economic rationale for such violations
(e.g. market failures, stimulation of R&D), we should not accept them.
But, there is no economic rationale.  There is not a iota of evidence that
swpat (with the unavoidable breadth) have any positive impact whatsoever
on the software economy.

The reason why swpat are so broad lies buried in the concept of technical
invention.  This is why insisting on this concept is so important.

By patenting uses of the universal logical device (the computer), you
patent logic itself.  The achievement for which you grant the patent is an
achievement of pure logic -- no iota of techncal character since the
technical aspect is all contained in the universal computer which is
already known.  Moreover the actions that your patent claims covers are
all actions of pure logic -- the universal computer is the standard
(most efficient and only relevant) means of expression of pure logic,
because it is its most abstract form of expression.

Patenting software is in every respect an outrageous idiocy, and any
government and any parliament that even discusses this possibility should
be deeply ashamed of itself.  A few decades ago people would have
been locked up for psychiatric treatment for proposing such a thing.

--
Hartmut Pilch, FFII & Eurolinux Alliance              tel. +49-89-12789608
Protecting Innovation against Patent Inflation       http://swpat.ffii.org/
120,000 signatures against software patents      http://www.noepatents.org/