From phm@a2e.de  Tue Apr  3 12:54:52 2001
From: phm@a2e.de (PILCH Hartmut)
Date: Tue, 3 Apr 2001 13:54:52 +0200 (CEST)
Subject: [Patents] basic choices
Message-ID: <Pine.LNX.4.33.0104031259470.9555-100000@mulix.oas.a2e.de>

> >evidently these unsolved problems are even more detrimental.
>
> But you are assuming that the patent system has to be "extended" to
> handle software patents, whereas I argue that software patenting is
> implicitly already assumed to be a part of the patent system.

We have the following basic choices

(0) no patent system (difficult in the short term)
(1) patenting of problem-solutions that contribute to the progress of
    natural science, i.e. provide new insights into the causal relations
    between using natural forces and achieving a useful effect
(2) patenting of "anything useful" / anything turing-compatible, i.e.
    any new idea that major economic players might like to own

Now there is dissent on

(a) whether options 1 and 2 are clear enough to be of any use
(b) whether the tradition of limiting patentability to "technical"
    things and not including math, software and business method among
    these corresponds to option 1 or is rather just based on errors
    and is really better represented by option 2.

Moreover there are some virtual debates about distinguishing between
"technical software" and "non-technical software" or "state-street"
vs a more conservative pro swpat approach.

Interestingly enough, even the patent maximalists the European Commission
and the British Patent Office, who are pushing for option 2, are
pretending to be advocating option 1 (i.e. "with technical effect only",
"no business methods" etc).

So in fact we have a general consensus that option 1 is wanted and that it
corresponds to the European tradition.

Moreover we have shown sufficient evidence to corrobate this.  Such as the
Kolle article of 1977, the BGH verdict "Dispositionsprogramm" of 1976, the
French verdicts of the 1970s, a long series of BPatG verdicts up to Aug
2000, the EPO and DPMA patent guidelines of the 1980s, the PatG
commentaries of Benkard etc.  Benkard, Kolle and others explain clearly,
why the non-patentability of software is implied in the technicity
criterion, which according to them is at the heart of a 100 year
customary law consensus and therefore not overturnable by any court
verdicts, even though the possibility of overturning it is, according
to Kolle, also implicitely present in patent law.

We have subjected option 1 (technicity criterion) to some random test
samples of current-day EPO patents and found it to be workable, leading to
a rejection of 3% of all EPO patents when applied in a cautious way and
possibly more, usually very trivial ones, when further extended.  Thus
option (1) can be fairly said to be

(1) the option meant by the current law and the only one compatible
    with it
(2) the sharpest of all patentability criteria that are present in
    European patent law today
(3) the criterion that more reliably than any other eliminates bad
    patents and that is best tested with respect to this

Our only problem is that we have a large group of lawyers who either never
really understood the technicity criterion or have in the late 80s or
early 90s decided to drop it by voiding the law of its meaning and vowing
to change it at the next opportunity (e.g. EU harmonisation).  Now it is
this group of patent lawyers that is in charge of decisions in Brussels.
And they have made sure that only members of their group can take part in
any lawmaking decisions, that dissenting law scholars are excluded and
that diverging caselaw is never mentioned.  While one of them, Pierre
Breese, is complaining about a "dictatorship of the free software lobby",
we in fact do not have a single representative in any of the
decisionmaking bodies.  And they are never even mentioning our regulation
proposal, which clearly formulates the meaning of the current law in
the sense of option 1

	http://swpat.ffii.org/stidi/eurili/
	http://petition.eurolinux.org/consultation/ec-consult.pdf

In America the problem is somewhat similar.  There may have been no
article like 52 of EPC that spoke clearly in favor of option 1.  Yet
option 1 has been present throughout the 60s to early 80s.  In Europe
options 1 and 2 were struggling in the 60s and 70s.  The BPatG actually
opted in favor of option 2 during that time and changed its mind only in
1975 at the time of the EPC signing, a year before the
Dispositionsprogramm verdict.  Which is why some patent lobbyists
correctly claim that software patents have been granted since the 50s and
60s in Europe.  Yet the change of mind in the early 70s was quite radical
and the rules laid out very clearly.  The problem is only that not all
people in the patent business were aligned on those new rules.  The schism
persisted and the losers of the 70s became the winners of the late 80s.
And it persists today as well, in Europe as well as in the US, where the
CAFC and the patent law community have dictated the rules in a similar
way.

It is wrong to say that there is any of the two above-stated options is
unscientific or untenable or that the answer is already implied in the
patent system.  It is of course correct to say that unlimited
patentability according to option 2 is implied in the current practise of
the EPO.

-phm









