From jp@smets.com Thu, 12 Aug 1999 13:19:12 +0200
Date: Thu, 12 Aug 1999 13:19:12 +0200
From: Jean-Paul Smets jp@smets.com
Subject: [Patents] Re: [Patents] Re: Patents] Answer to requests sent to french Economic
Content-Type: multipart/mixed;
  boundary="------------1B997DC262D81213B4B9C741"

This is a multi-part message in MIME format.
--------------1B997DC262D81213B4B9C741
Content-Type: text/plain; charset=iso-8859-1
Content-Transfer-Encoding: quoted-printable

Nick Fleming wrote:

> One point I would like some input on is, assuming the
> worst, what limits should be placed on software
> patents ? =


Here is a short answer.

First, there are two kinds of software patents.
	A- patents on inventions which include some software process (legal in
Europe)
	B- patents on computer programmes as such (illegal in Europe)
This distinction is extremely important. A computer programme as such is
not considered in Europe as a potential invention with industrial
application.

A- Patents on inventions which include some software process

They are fully legal in Europe (and US and Japan).

Example: patent on a process which includes some software and allows to
discover oil under the earth
Example: patent on a process which includes some software and allows to
decompress digitized music and listen to it

The EPO are the guys who decide the limit to what can be patented or
not. Simply because patents are usually written by patent advisors who
are former patent reviewers at the EPO and are better paid than what the
EPO can, there is a trend to accept more and more patents on more and
more stupid processes. This trend is very hard to stop because
	- the well-paid best experienced guys are paid to go through the
reviewing process managed by the low-paid least experienced guys
	- it is hard to define in terms of law what a "stupid process" is

The main legal approach that one can think of is to ask the EPO to
restrict itself seriously to inventions with the following definition:
	a- new
	b- inventive step
	c- industrial application
Point c- is the most important, since it allows to reject many software
patents if applied properly and strictly. But the EPO is trying to push
a new definition of inventions:
	a- new
	b- technical solution of a technical problem
Which allow to patent many more things. This new definition should be
fought strongly and people should be reminded article 52 (see part D-)

As a reminder, in the US, the legal definition of inventions is:
	a- new
	b- non obvious
	c- useful
which allows to patent about anything.

The current situation in Europe has a big advantage for software
publishers. In Europe, a software can not be considered today to
infringe on a patent. Only the whole industrial process involving
software can be considered as eventually infringing. Software itself is
only an sub-part of the whole patented process. Therefore, software
publishers can only be sued for providing some sub-part which enables
their custommers to assemble a patented process. This can only be done
by notifying them first that they are providing some sub-part of a
larger patented process whereas suing someone for direct infringement
can be done without notice. It gives some kind of protection to software
publishers against legal terrorism.

It is very important to keep this protection. Otherwise, someone who
publishes free software on the net could be sued without notice and lose
his house. The way I recommend to go is to insist on two different
rights
	1- the right to copy a software (copyright, copyleft)
	2- the right to use a software (end users should pay patent licenses
which are independent of the software they use, free or proprietary)
Those two rights should be kept independent. I am currently writing a
paper to explain why in terms of economy and how it can be implemented
in practice.

In Europe and Japan, patent should not be legally enforced on CVS
archives because CVS archives of free software can be considered as
virtual R&D centers and patents are not enforceable for research
activities. This does not apply to US since in the US, even researchers
need to ask the patent owner's permission to use a patented process
within a larger research project.

B- Patents on computer programmes as such

Illegal in Europe. Legal in the US.

This is what we really do not want in Europe because
	1- software publishers could be sued for infringement (and all of them
are already infringing on some stupid patent without knowing)
	2- it would become possible to patent social practices and business
methods implemented by a computer programme

Point 2- is the most important. Currently, a new voting system based on
the web, a new automatic FAQ system based on the web can not be patented
in Europe because they have no industrial application. But if programmes
as such can be considered as inventions and if inventions are only
"technical solutions of technical problems", then it will be feasable
within a few years to get patents for such social practices of business
methods implemented through software. This is very dangerous for the
future of the Internet, the information society and business.

There is also a legal reason to reject such patents. Any computer
programme is scientifically equivalent to a maths proof which can not be
patented as such. Allowing patents on programmes as such in Europe would
lead to a situation where the same object would be patentable and non
patentable.

Of course, this does not apply to the US where maths proof can
eventually be patented.

C- Comments on the response from various French authorities

There are two different answers
	1- the one from the French State Secretary of Industry who writes
something like "the government cares about possible negative impact of
software patents and especially on free software"
		=

	Des inqui=E9tudes se sont cependant manifest=E9es r=E9cemment quant =E0
l'impact
	qu'aurait cette mesure sur les entreprises de logiciel, et notamment
	pour les logiciels libres. Le Gouvernement a pris note de ces remarques
	qu'il =E9tudie avec attention. =


	2- the one from the General Directorate of Industry, Information
Technology and Post (DIGITIP) which is the same as the one from the
French Patent Office (INPI) which does say anything like that. They
probably do not care about free software and do not believe software
patents may have negative economic impact.

The letter from the French State Secretary of Industry says that
"inventions are considered by the EPO as technical solutions of
technical problems" whereas article 52 states that "European patents
shall be granted for any inventions which are susceptible of industrial
application, which are new and which involve an inventive step."

	La pratique d=E9velopp=E9e par l'OEB consid=E8re une
	invention comme brevetable si elle pr=E9sente une "contribution
technique"
	=E0 l'=E9tat de la technique.

A new question should be therefore sent to the French State Secretary of
Industry asking him some clarification. Something like "why does the EPO
consider  technical solutions of technical problems to be inventions and
forgets about the industrial application requirement clearly stated in
the M=FCchen convention as well as in the French Law".

The introduction of the letter from the DIGITIP/INPI is somehow
misleading because it says "computer programmes can not be patented in
Europe while software can be patented in the US" which could be
understood as "software can not be patented in Europe while it can be in
the US" which whould be wrong.

	L'article 52 paragraphe 2 de la Convention de Munich et les lois
	nationales des Etats membres de l'OEB ne permettent pas la
brevetabilit=E9
	des programmes d'ordinateurs en tant que tels. Aux Etats-Unis et au
	Japon, en revanche, les logiciels sont clairement prot=E9geables par
	brevet. Il est n=E9anmoins possible de breveter en Europe une invention
	technique qui fait appel =E0 un logiciel.

The introduction of the letter from the French State Secretary of
Industry is much more precise, clearly explaing the difference between
software related patents (legal in Europe) and patents on programmes as
such (illegal in Europe).

	L'article 52 paragraphe 2 de la convention de Munich et les lois
	nationales des Etats membres de l'office europ=E9en des brevets (OEB) ne=

	permettent pas la brevetabilit=E9 des programmes d'ordinateurs en tant
que
	tels. Toutefois, il est possible en l'=E9tat actuel du droit de breveter=

	une invention technique qui fait appel =E0 un logiciel. Il existe ainsi
	environ 13 000 brevets europ=E9ens portant sur du logiciel. Aux
Etats-Unis
	et au Japon, les logiciels sont clairement prot=E9geables par brevet. =

	Cette situation semble insatisfaisante.

Both letters include a very misleading mistake when they talk about
"removing software from the list of non patentable inventions in article
52.2". What 52.2 is about is not software but computer programmes. What
52.2 says is that programs for computers as such can not be considered
as inventions. But industrial processes based on software are already
patentable in Europe as 52.3 states (see section A- of this email).
	=

French State Secretary of Industry :

	la suppression des logiciels de la liste des inventions non brevetables
de
	l'article 52 paragraphe 2 est apparue utile aux intervenants. =


INPI/DIGITIP

	La suppression des logiciels de la liste des inventions non brevetables
de
	l'article 52 paragraphe 2 est apparue =E0 cette occasion tr=E8s
	consensuelle, notamment lors de l'audition organis=E9e =E0 Luxembourg le=
s
25
	et 26 novembre 1997. =


Moreover, INPI/DIGITIP talk about "consensus" when the French State
Secretary of Industry says "it might be useful".

Bothe letter include a surprising hidden assertion "software patents
might be useful for small companies"

	Un tel manque de transparence est dommageable car il =E9carte de fait du=

	syst=E8me une majorit=E9 des PME du secteur des logiciels : celles-ci
	ignorent qu'il est possible d'obtenir une protection par brevet

Conclusion:

	- INPI wants software patents and does not care about economic impact
or free software (of course, more patents means more budget for them)
	- DIGITIP wants software patents and does not care about economic
impact or free software (they only listen to big players including
Bull/IBM/Microsoft)
	- French State Secretary of Industry has a balanced position

For your information, another anwer was received from the directorate
for SMEs which was much closer to the position of the French State
Secretary of Industry.


D- Article 52 of the M=FCnchen convention

(1) * European patents shall be granted for any inventions which are
susceptible of
industrial application, which are new and which involve an inventive
step. =

(2) The following in particular shall not be regarded as inventions
within the meaning of
paragraph 1: =

(a) discoveries, scientific theories and mathematical methods; =

(b) aesthetic creations; =

(c) schemes, rules and methods for performing mental acts, playing games
or doing
business, and programs for computers; =

(d) presentations of information. =

(3) The provisions of paragraph 2 shall exclude patentability of the
subject-matter or
activities referred to in that provision only to the extent to which a
European patent
application or European patent relates to such subject-matter or
activities as such. =

(4) * Methods for treatment of the human or animal body by surgery or
therapy and
diagnostic methods practised on the human or animal body shall not be
regarded as
inventions which are susceptible of industrial application within the
meaning of paragraph 1.
This provision shall not apply to products, in particular substances or
compositions, for use
in any of these methods.
--------------1B997DC262D81213B4B9C741
Content-Type: text/x-vcard; charset=iso-8859-1;
 name="jp.vcf"
Content-Transfer-Encoding: base64
Content-Description: Card for Jean-Paul Smets
Content-Disposition: attachment;
 filename="jp.vcf"

YmVnaW46dmNhcmQgCm46U21ldHM7SmVhbi1QYXVsCnRlbDtjZWxsOiszMy0oMCk2NjIgMDUg
NzYgMTQKeC1tb3ppbGxhLWh0bWw6RkFMU0UKdXJsOnd3dy5zbWV0cy5jb20KYWRyOjs7Ozs7
Owp2ZXJzaW9uOjIuMQplbWFpbDtpbnRlcm5ldDpqcEBzbWV0cy5jb20KeC1tb3ppbGxhLWNw
dDo7MApmbjpKZWFuLVBhdWwgU21ldHMKZW5kOnZjYXJkCg==
--------------1B997DC262D81213B4B9C741--





