[Patents] Re: Hardware/software codesign

[Edward Avis epa98@doc.ic.ac.uk]

>From: Gregory Aharonian <srctran@world.std.com>
>Subject: RE: [Patents] Hardware/software codesign

>There is nothing special about software,

Well... that's certainly a matter of opinion.  Software (which is just
information, like a novel or an audio recording) is certainly very
different from physical devices.  But to continue:

>it is the
>latest area of technology to be "hot" and therefore stress the
>patent systems around the world.  All of these arguments on this
>list have been heard repeatedly for the last one hundred years
>(i.e. "why should transistor circuits be patentable - they are
>all obvious from early vacuum tube circuit patents"),

I'm fairly confident that the situation for software is much worse than
it ever was for those new areas of patentability.  And computer software
has been around for several decades, in the US it has been patentable
for quite a large chunk of that time.  You would expect, if the system
were going to settle down and reach a sensible balance, it would have
done so by now.  Yet things are steadily getting worse.

>In fact, in terms of disruptions
>to research and development, suffering by consumers, and
>distortions in the marketplace, biotechnology patenting is
>causing orders more problems than software patents (i.e. if
>you can't afford access to my patented software product, you
>go read a book - if you can't afford access to my patented AIDS
>drug, you die).

True, but the costs of R&D are so high that the choice may be between a
patented AIDS drug and no drug at all (since companies will not bother
investing without sufficient incentive).  I see no evidence at all that
this is true of software where the hard work of development would
proceed with or without patents, which are normally used as an
afterthought by the legal department to get leverage with competitors.

Also, the effect on development is much worse, because there are many
small developers and individuals in the software market, which isn't the
case for pharmaceuticals.  A big drug manufacturer can afford to
maintain its own collection of patent lawyers and negotiate with a small
number of patent-holding companies; compared to the startup costs and
costs of R&D, it's not too bad.  But for an individual developer or SME
under threat from thousands of software patents (even if they turn out
to be invalid), it's just too much of a burden.

The effect on users is also quite serious, though of course it's nothing
close to unaffordable AIDS drugs.  In the drugs market, the effect of
patents is to increase prices for consumers; but in software, it
directly reduces choice since companies are scared away from entering
particular market segments.

(Patentability of drugs does not reduce choice in this way because there
is not much to choose between two rival brands of exactly the same
chemical.  Whereas there is certainly a benefit in having many competing
products performing the function of MP3 player, or CSS-capable web
browser.)

>I have no problem questioning the patentability of software, as
>long as it is done in the context of questioning the patentability
>of all areas of technology as we enter the 21st century.

I'm knowledgeable about software but not biotech, so I'll stick to
one particular issue.

>But there is no science and no logic in separating out software.

The patent system is not created on the basis of 'science' nor on a
patent lawyer's idea of what is logical.  It exists only for its
benefit to society.

-- 
Ed Avis
epa98@doc.ic.ac.uk