[Patents] Economic assessment

[PILCH Hartmut phm@a2e.de]

Ed Avis responded to an often-reiterated troll question:

> >When it comes to
> >patent law, I dislike the uses of the words "hardware" and "software",
> >and prefer "computing" or "automata" patents, where the functionality
> >(and its novelty and nonobviousness) is the only thing that matters,
> >and its expression (hardware, software, optics, etc.) is a marketing
> >decision.
> 
> Certainly there are some things that can be implemented both in hardware
> and in software.  But that doesn't mean that the expression is
> irrelevant or that it is hard to determine.

You response is good, but the question was wrong anyway.
As often repeated:  we are not proposing a defintion of the limits of
patentability by the concept of software vs hardware but rather of
technicity / physic[al]ity / transformation of matter, the result of which
is that software generally does not qualify for patentability. 
Putting an extra disclaimer for computer programs into Art 52.2 is just
an extra layer of security against misinterpretations of the law.
The lawmakers of 1973 really did a very good job.

Here's another version of the troll question:
 
> >People on this list might want to try to come up with a definition of
> >"software" that can actually be used to debate these issues in Europe.
> 
> How about this: pure information whether in printed or electronic form
> does not infringe on any patent.  If a computer system is noninfringing
> when considered without any software running on it, it is also
> noninfringing when running software.  Software is defined as information
> (whether program code or just data) which is input into a computer.

and another one
 
> >Most of the American definitions of the terms encompass non-"software"
> >concepts such as digital hardware circuit diagrams and other electrical
> >and mechanical systems.
> 
> Of course a circuit diagram should not infringe any patent - otherwise
> the patent document itself would be illegal and so would some
> electronics textbooks.  But manufactured circuitry - something you can
> hold in your hand - which is based on a circuit diagram may infringe on
> a patent.  (Although it isn't the particular circuit but any circuit
> using the patented techniques.)

... where the patented techique consists in making use of newly found
relations of physical causality, not in using obvious means for
implementing a new algorithm.  A pure algorithm can never be patented, not
even through circuitry.  As has been explained very well in Kolle's
article of 1977.

-phm