[Patents] Dutch MP's demand triviality criteria in the law before allowing SW pats

[Gregory Aharonian srctran@world.std.com]

>In fact, you appear to agree with me that the patenting an instruction
>set is for avoiding competition. It is not used to protect research.

Patents protect research by preventing competition.  All Intel can get
a patent on is any new and unobvious features to their instruction set.
Those parts of the instruction set already in the public domain before
the patent was filed remain in the public domain.  If the PTO allowed
Intel to claim features of the instruction set already in the public
domain, than that is not an indictment of the patent system or Intel's
strategy, but rather another example of quality problems with examining
patents.

With regards to RISC, it it loosing out not because of patent practices,
but rather CISC got fast enough to make a RISC transition too costly,
and clusters of CISC using software like PVM available on the Internet
allows using to handle computationally intensive calculations with
bunches of cheaper CISC.

Greg Aharonian