From phm@a2e.de  Fri Dec 29 18:21:05 2000
Date: Fri, 29 Dec 2000 19:21:05 +0100 (CET)
From: PILCH Hartmut phm@a2e.de
Subject: [Patents] patent pools enforcing per-copy-licensing mode of
 software  development

> Hello, Mr. Garcia Fernandez,
>=20
>    as my departure is delayed a bit, here a short answer:
>=20
> Ram'on Garc'ia Fern'andez schrieb:
>=20
> > Do you believe that a patent should be granted in exchange for any
> > idea (method), or only when there is a costly research behind?
>=20
> If the effort to achieve an invention is the only measure, only expensive=
 and
> non-effective
> research and development will be honored.
>=20
> > What is your opinion about trivial/easy patents? What measures would
> > you suggest against them?
=20
> In Germany, we have a criterium called "Erfindungsh=F6he" (meaning
> significance of the invention).=20

No, we don't have that criterion any longer in the official vocabulary.
We have only "erfinderische Taetigkeit", i.e. "inventive step".=20
Also, "Erfindungshoehe" means "height of invention", i.e. the distance of
the technical level of the invention from prior art.  You could also call
it "depth of invention" (Erfindungstiefe), meaning the same.
This criterion has been given up, as it is not workable.
Or in what unit do you want to measure it?  cm?  twists of brain?

Another criterion that was valid until the 70s and was then given up is
called "contribution to technical progress".
Actually the BGH (Federal Court) recently ruled that this is one reason
why software must be patentable:  it is no longer required that any
contribution to technical progress be made so therefore a computer that,
in terms of technical contribution, does not go beyond the prior art,
must still be considered for patenting.

All the nice virtues of the German patent system have been gradually
dismanteld, and software patents are just the latest consequence of this.
There introduction is a direct consequense of the lowering of requirements
for inventivity.

> This is an important yardstick and requires professional knowledge to
> apply correctly. If not applied, we would indeed be flooded by trivial
> patents..

We *are* flooded by trivial patents, and you would know it, if you
wanted to know it.  Look at=20

=09Horror Gallery of European Software Patents
=09http://swpat.ffii.org/vreji/pikta/

=09Why Software Patents are So Trivial
=09http://swpat.ffii.org/stidi/frili


=20
> > Often patented inventions are accidentaly reinvented and infringed.
> > In those cases, patents are not helping the society at all: we
> > would learn the invention anyway, perhaps some months later.
> > What do you think about that?
>=20
> If I violate a law that I did not know about, my ignorance is not accepte=
d
> by the courts as an accuse. For this reason, I have been arguing that all
> juridical literature should be "public good" and free of charge. I have n=
ot
> gotten much support yet. I am also asking that knowledge about patented
> inventions should become a significant part of basic engineering educatio=
n
> (not just in software engineering). As you know, patents are extremely
> well documented, easily accessible and freely useable for education and
> research.

That is not an answer to what Ramon asked.
=20
> > > True software engineering, he says, needs patents, and the hobbyists
> > > should be sacrificed for that.
> >
> > Software patents began to be used in the US after 1982. (Before that da=
te,
> > they where occasional and very rarely enforced). The most important
> > development of the software industry where before 1982. Facts apparentl=
y
> > contradict your statement "True software engineering needs patents".
>=20
> The best know German software patent was issued to F.L. Bauer and
> his colleage Samelson in 1957.=20

There were a few cases in the 50s and 60s that were later invalidated by
decisions of the courts (BPatG and BGH).  German courts were
drifting between two viewpoints at that time, and the definite
clarification came with

(1) the EPC
(2) the Dispositionsprogramm decision of 1976, preceded by a minor
    decision of the BPatG in the same direction

> Bauer's name became later associated with the term "software
> engineering". As I said in a previous note, I think that the
> difference between science and engineering can be explained by the
> difference between discoveries and inventions.

This is just one criterion.  The debate is also one of the 60s, and you
can find it well analysed in Kolle's articles of the 70s.

Your arguments apparently haven't changed since then.  They can still be
refuted with the same counter-arguments as were used then.

> To make the long story short: If software development likes to become
> engineering, they cannot be satisfied with discoveries (as the ones
> made by Pythogaras, Mendel and Maxwell), but rather go for inventions
> (like Edison, Marconi, etc). To demonstrate that one's inventions are
> "new, useful and non-obvious" the seal of approval of a patent
> examiner can be more helpful than the acceptence by a journal or a
> conference. I have been an editor-in-chief of a jounal for 15 years
> and know how tempted one can get if one needs to fill an issue. That
> is all, I have been saying.

That software innovations have some things in common with "inventions" as
opposed to "discoveries" is an argument for making them patentable.  But
not a sufficient one.

> One final remark for Bernard Lang: I appreciate and share his concern
> for individual freedom. But if scary people demand that police be
> visible in cities and on highways, this certainly limits the freedom
> of normal people who want to draw money from banks and want to race on
> the German Autobahn. Maybe the police catches some robbers or drunken
> drivers also.

All Bernard asks for is that you explain the social benefit that can be
achieved by limiting freedom of expression.

You have successfully argued the case for a police state:  it creates a=20
social benefit which consists in more security.  Therefore we can then
procede to weigh that benefit against the disadvantages of a police state.

However until now you have failed to argue the case for software patents. =
=20
The only social benefit you mention is a theoretical one, i.e. the
incitation to disclose ideas.  But you have not made it plausible that
software patents have in reality led to the disclosure of valuable ideas
which would not have been disclosed otherwise.  Moreover you have argued
that opensource software must be sacrificed for the patent system, with
the effect that the disclosure and diffusion of ideas is slowed down.  So
on the issue of disclosure you have effectifely shown that software
patents have no merit for society.  And you have not shown that they have
any merit on any other issue.  So it seems that the case for software
patents is even harder to argue than the case for the police state.

-phm





