[Patents] English translation of SSLUG.DK's comments on the EPO Basic Proposal

[PILCH Hartmut phm@a2e.de]

Here are some small criticisms
=20
> Do not forget that the EU should not be a cooperation of
> wealthy companies, but should be the European Union of the

I=B4d rather say "cooperation of patent professionals".
It is not at all clear whether wealthy companies want or need software
patents.

> Before 1994 companies were very successful without having to
> obtain and use software patents. Microsoft is an example of a
> successful company, that has made its progress primarily
> without the usage of patents.

I'd mention a few more companies, because patent lawyers like to claim
that Microsoft acceded to monopoly power only because smaller companies
were not protected by patents.  Maybe cite the opposition of Adobe and
Oracle to swpats.
=20
> The rationale behind patents is to protect the inventor,
> while furthering innovation and development by making
> the inventions public after a certain period of time.
> The question whether software patents further innovation
> has mostly been examined by economists, and the inquiries
> shows that software patents, at best, have no positive
> effect on development.
=20
> EU - IPI-report:

There are quite a few more, see http://www.pro-innovation.org or the
recensions in the appendix of the Public Letter
http://swpat.ffii.org/xatra/epue28/epue28de.pdf. The former is much more
worthy of being called an economic study than is the IPI report.
=20
> What does the society gain by issuing patents?

The question "what does an innovative software developper gain?" should be
asked first.  Actually he gains just as much as a composer of music would
have gained if a well-meaning Austrian government had instituted in 1770=20
a system for exclusion rights on composition techniques.  The difficulty
of writing a piece of software is similar to the difficulty of writing a
big symphony.  It is well protected by copyright.  And, unlike with music,
plagiarism is extremely difficult, because decompilation doesn't yield a
useful result.  Any imitator has to do the whole work over again and there
is hardly any free-ride effect.  Any such effect is outweighed by the
"first mover takes all" effect.

This is extremely important, because non-programmers don't know it.  It
should become before any mention of economic analyses.
=20
> Originally "technical" had a very narrow definition, formulated
> by Professor Kolle, a recognized legislative author, more than

not Professor, as I stated before.
Actually, Kolle is today the head of the EPO's International Relations
department.  And pro software patents at least in his function as a
representative of the EPO.
So it would be better to quote the German Federal High Court
(Bundesgerichtshof BGH) which gave precise definitions for technicity and
its application to software in the "Dispositionsprogramm" decision of
1976, which was then analysed by Gert Kolle in an excellent and profound
article in GRUR of 1977

=09http://swpat.ffii.org/vreji/papri/grur-kolle77de.html

The URL should be quoted.  Moreover an excellent new article

=09http://swpat.ffii.org/vreji/prina/patpruef.pdf

by a software examiner of the German Patent Office should be quoted as
well.  And of course don't forget the decision of the Federal Patent Court
BPatG

=09http://swpat.ffii.org/vreji/papri/bpatg17w6998de.html

which restated in August 2000 that computer programs may not be claimed
and are not technical.

> If the current exception of software patents is removed from
> Article 52 in the M=FCnchen Convention, then all software can be
> considered "technical" consequently the word "technical" will
> lose its meaning. In that case everything can be regarded
> as being technical, the result will be a patent system without
> limits as in the USA. A technical knitting pattern or a
> technical baking recipe becomes possible.

You should mention some EPO patents, which are in fact already as
untechnical as those in the US, see our horror gallery

=09http://swpat.ffii.org/vreji/pikta/
=20
> The IPI report mentions different scenarios. It regards the
> translation of the existing EPO practice into law as being
> status quo. However, this is a mistake, as in many instances
> the current EPO practice is in conflict with the national
> practice. And national patent practice takes precedence over
> EPO practice as this is mere jurisprudence not law as the
> national practice.

see BPatG example
=20
> Patent offices, on the other hand, describe an idea or invention
> in patent legalese, which makes it impossible for even experienced
> programmers to recognize an idea from the written description.

not quite impossible.
the claims are in legalese.
the descriptions are not really in legalese.
but still hard to read and without reference implementations.
=20
> Nor is it possible to precisely define the broadness of
> applicability of a software patent, as a particular algorithm
> can be used in many different applications. This make software
> patents very different from traditional patents.

I am not sure whether this is a real difference.
Claiming such things makes you vulnerable.
=20
> Software is special as the idea behind the software often
> carries information. This is particular for software
> inventions and this forces us to assess the impact on the

never use the word "invention" in the context of software.
Actually even "software innovations" must be used with care.
I'd rather say "new software problems" because that's what
EPO software patents are about.  They don't even patent the
solutions but the problems themselves.  See our horror gallery
and the recommended article by the German patent examiner.

> development of society, when a software invention is

no "software inventions"

> patented. The incremental development of society is impeded,
> if the patent of less advanced software, effectively prevents

"patents have nothing to do with software"
as Eitel Dignatz said.
Also, don't create the impression that the problem is novelty.
The patent people are quick to claim it can be solved.
Solving it doesn't help anyway.  Of course lots of new software
problems come up every day.  Nevertheless they mustn't be patented.

> the development and application of better and more effective
> software.
=20
> We would also like to point out that software is a
> combination of technique and representation of information.
> It is not possible to distinguish between the intellectual
> and technical aspects of a computer program. To allow

There is no "technical aspect of software", at least in the sense in which
we want to define "technical". =20

What about mentioning the following definition of invention:

An invention is an embodiment of a teaching on how to use natural forces
in a new way to directly cause a non-obvious result serving the purpose of
industrial production of material goods.

This could be directly used in Art 52, if we were to make it really clear.

=09http://swpat.ffii.org/stidi/epue52/

In the USA, the traditional definition until 1980 was that an invention
must be a new way of "transforming matter", which is quite the same.
 =20
> patents for a mixture of method and information would
> be charting a course into a legal minefield.

This seems a bit unclear.  What is a legal minefield?
=20
-phm