From phm@a2e.de  Sat Oct  7 19:18:34 2000
Date: Sat, 7 Oct 2000 21:18:34 +0200 (CEST)
From: PILCH Hartmut phm@a2e.de
Subject: [Patents] Re: french views

> Stefane,
>     Try reading what I wrote.  As I said, in the early 1990s, Oracle
> did have a position against software patents, and circa 1994 wrote
> that statement you refer to at the lpf.ai.mit.edu Web site.
>     But shortly thereafter, they abandonen those beliefs, and started
> an active software patenting campaign.  So your silly question "Did
> they write something that said they changed their mind?  I believe
> they didn't" is pointless - their actions speak much louder than 
> their words - they are actively patenting software, engaging in the
> same questionable practices that most other companies patenting
> software engage in.

It is only a matter of normal rational behavior that you have to
patent everything you can, if this is the rule in society.

If we lived in a state of war, we'd all be carrying guns and shooting at
each other.  This wouldn't allow any conclusions about our beliefs.

Thus again your conclusions are not supported by your evidence.
 
>     Now Hartmut tries to defend Oracle's switch in attitude by trying
> to argue that Oracle still may be against software patents, but is
> patenting because everyone else is.  

I did not say that Oracle is against software patents.  I didn't even say
that Oracle ever was against software patents.  I just disputed that there
is any evidence for a change of attitude.

> Still false.  With Oracle's clout
> and resources, there is much the company could do to help make the
> software patent examination process at the USPTO more efficient and
> of a higher quality - Oracle could help fight the abuses of the current
> patent system.  

Again this would be irrational behavior.  Oracle is a big company that has
to make money, and the patent system has something to offer for such a big
company.  It allows them to fence in their territory and keep most other
companies out of the database business.  The worse the examination process
the better the chances for Oracle to monopolise it.

On the other hands, even if Oracle wanted to invest ressources into
efforts to improve the patent examination system, such ressources wouldn't
really help anybody.

>     None of the weak arguments I have seen here recently have changed
> my long held belief that the solution to the problem of bad (software)
> patents is to get rid of the bad (software) patents, not to get rid of
> patents in general (which is what some of you are arguing, though
> hiding with the same type of word games that lawyers play).

Again you evade my questions and fail to substantiate your position with
arguments.

Do you mean that reducing the patent system to real inventions (i.e. new
ways of using natural forces to solve a material problem) means getting
rid of it in general?

I don't think so.

There are still lots of real inventions going on.  I am dealing with
several real invention patents every month.
New magnetic materials are usually not found by pure reason but by trial
and error and intelligent + costly experimenation.
Man has not yet arrived at a stage where all of nature is
simulatable and predictable.  I guess we will never arrive there.  But the
legitimate realm of the patent system could gradually become narrower,
because once nature is simulatable, there is no longer any real invention.
Eventually everything could be invented by simulating robots.

Let's go one step back and assume my distinction between a logical and a
physical world was not workable.  Ok, Greg, is your approach to improve
the patent system by finding prior art actually workable?  I'd say no.
Many very harmful patents do contain something new that can't be refuted
by prior art.  Most of these patented logical innovations seem obvious in
hindsight and it would take most people very little time to find them.
Often thousands of people will find them in less than 1 hour once they
stumble across the problem.  Sometimes even robots will find them.  Thus
you need a way for sorting these out as "obvious".  But the patent system
doesn't offer such a way.  Do you know one?  Now going even one step
further back.  Assuming you had a way of tackling the obviousness problem.
Some very rare software patents are new and non-obvious and even involve a
fair amount of work.  But yet they cause great harm, e.g. by blocking a
standard or by being uncircumventable and making it impossible to solve an
important task under opensource conditions -- "the best things in life are
free" was said by a famous American judge in the 70s in the context of
software patentability.

Since you don't try to provide answers to any of these problems and
concentrate only on the technicality of finding prior art, at which you
are an expert with proprietary offerings of high market value, I must
doubt whether you are genuinely interested in helping to solve the
problems caused by the patent system.  The worse these problems get, the
higher the market value of your proprietary offering will be.  Any
solution to the core problems of inventivity or physical substance is a
threat to this value.

Apropos "physical substance":  I found this term in Japanese literature on
software patents, together with the term of "concreteness".  The Japanese
patent system has the same concepts as European jurisdiction has
developped, and even when software patents were introduced in Japan, they
were explicitely limited to software solutions involving the use of
natural forces.  But all this was diluted under American pressure and
because of the fact that there is no democratic counterweight to the JPO,
which within a weakened and less self-consious MITI has gradually 
made itself independent.  

>From my experience in talking with Japanese business people, I am quite
confident that if Europe can return to its fine patent law tradition,
Japan will very soon follow suit.

-phm






