|
Authored by: Anonymous on Monday, July 23 2012 @ 01:34 PM EDT |
Yes, this tangent has basically nothing to do with "process
patents".
It might be relevant to whether an English (or "patent-ese")
description of a process is really a correct description,
and therefore whether that description can possibly
constitute "constructive reduction to practice", since it's
pretty easy to describe something in English that's
impossible for a program (or a human, for that matter) to
do. But for Groklaw purposes it's not patenting the
impossible that we're worried about, it's patents granted on
stuff that is also (previously or very soon afterward)
independently implemented without much effort or research,
i.e., empirically too obvious to be patent-worthy.
My self-appointed role in this discussion is to describe the
perspective of a judge or other non-computer expert, who
thinks that the "system" (including software) does new
things, so why *not* patent the system? The argument that
CPUs are stupid (can't understand semantics, for one thing)
just goes to show that the program must be very smart.
Doesn't that make the program even more deserving of patent
protection? (Note the unfortunate effect of the habitual
language used by patent lawyers. Patent "protection"!)
Even judges realize that a computer program in combination
with general-purpose hardware isn't exactly a "new machine"
as "machine" was previously understood - but they think it's
equivalent, that new technology has created a new class of
inventions that might as well be called "machines" (after
all, even computer programmers use "machine" in ways that
include software), and they're struggling to find a way to
fit their limited knowledge of the new thing into the law's
old conceptions of "machine". So far, as PolR points out,
they're failing. But by the time they pay attention to
PolR, they'll be able to turn to a dictionary and find a new
definition of "machine" that fits their needs.[ Reply to This | Parent | # ]
|
|
|
|
|