I think there should be a kind of compensation
for using research
investments of
others. But that can not be by the actual patent system. A
link
to the invested
amount is needed and is missing in that patent system.
I'm not arguing that the patent system is fit for purpose
-
only that its originally intended purpose is actually a
good one and we need
something to fill that gap.
The patent system speaks of
inventions. But
every one making the same research
effort will probably come to
the same or similar results.
What invention?
I don't think
that's necessarily true. The MP3 example
is not just conducting a bunch of
experiments and then
seeing the result fall out of the results - its combining
what was learnt in that research with some good insights
into applied math and
signal processing and a good
understanding of what's possible on current
computational
devices to produce something that is (was at the time) both
truly innovative and a tremendous public benefit. But even
if any suitably
competent researcher were to eventually come
to the same answer, the purpose of
the reward is still to
encourage them to put the effort in to get to that
answer
and I think that is still worthwhile.
As a software engineer, I
completely disagree with the
crowd that argues that software is mathematics and
should
not be patentable. What I do is every bit as creative and
innovative
as what a mechanical engineer or an electronics
engineer does. I make complex,
useful structures out of
fundamental building blocks in exactly the same way
they do.
Why should innovation in software be less rewarded than
innovation
in hardware?
For the most part, what I do is fairly obvious and
similar
to what other people have done and copyright seems
to be the mechanism for
rewarding that effort that best
fits. Just as most engines designed by a
mechanical
engineer are pretty similar to other
engines and most buildings
designed by a structural engineer
are like most other buildings
and most
integrated circuits designed by electronic
engineers are pretty much like other
integrated circuits. But when someone devises something
truly new and
creative and innovative, I think a patent is the right
answer, irrespective of
whether it's hardware or software.
Again, the problem is not software patents,
but that the
quality of patents (and particularly software patents) is
abysmally low. Most of what they
cover is just mind-bogglingly obvious. But
every now and
then you come across an idea in software that leaves you
thinking, "Oh, that's cool. I would never have thought of
that in a million
years." Those ideas should be rewarded.
Suppose the
results of a research effort of a
first company are in practise kept
secret.
Hidden in vague, incomplete patent language. Hidden
in the total mass
of
patents. And somebody had to make the same effort giving
similar results.
There
is no reason he should pay that first company. He did not
use their
research.
He should not have to be afraid of getting a visit from a
company, asking money
for something they found out. Like he should not be
afraid
of a visit from the
mafia, asking money for what he already
owns.
No way a patent system can function without simplicity
and
clarity. And that is
completely missing know.
Again, I'm
not arguing that the patent system is fit for
purpose. But the idea behind it
is a good one, and it's
meant to solve exactly this problem. Without a patent
system or something like it, any invention is naturally kept
secret as the
only way of profiting from it. With a patent
system, we say that we'll let you
exclusively use your
patent for a fixed time in return for making it public so
that people don't have to repeat your effort. It should
make it easier for
the practitioner of
a discipline to keep up with the state of the art of that
discipline. Obviously
that's not the effect the patent system has today, but
the
idea is a good one. The problem is that the bar is set too
low for
obtaining a patent and too high to defend against
one, and this makes gaming
the system more profitable than
following its intent. [ Reply to This | Parent | # ]
| 
