[Patents] BPatG caselaw

[PILCH Hartmut phm@a2e.de]

> Seriously though, at some point we will need a fully informed debate
> in Europe about the pros and cons of extending the field of patentable
> subject matter.  Should patents be granted for computer programs as
> such?  Should patents be granted for methods of doing business?  The
> US has already made its choice on both of these matters, 

Even there, people complain that this choice was made by courts such as
the CAFC without participation of the lawmaker, see the "Digital
Millenium" report and the viewpoints expressed by Lessig and O'Reilly in
public debates.  Recently, with the Berman act an attempt was made to
bring about a debate about business method patenting to Congress.

> but there is a danger that any consultation exercise will misfire over
> here because of a misunderstanding about what is patentable now.

That can be kept separate.
A consensus on the interpretation of the EPC is not a prerequesite for
discussing what rules we should formulate for the future.
 
> I would be interested to learn more about the Federal Patent Court
> decision to which you refer.  

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

> You say that the Court rejected claims to a "computer program" or
> "computer program product" for not being acceptable as statutory
> subject matter under Art 52.2c.  Did the Court, at the same time,
> accept a claim to a (conventional) computer running the program?  

No. The proceedings did not discuss this matter.

The court stated only in general terms that for process running on a
computer to be a technical invention, this process must provide some new
means of making use of natural forces.

In fact it might have been easy to destroy the IBM patent in question even
on terms of novelty, but that was not discussed.

> In other words, is this just an example of form being exalted over
> substance?

The IBM attorney tried to portray it this way, saying afterwards to
the press that it "paradox" that the process could be granted but not the
program.

Of course this is nonsense, because the proceedings, as said above, did
not discuss whether the process was patentable.

Even if the BPatG had found the IBM's process of "searching erroneous
strings in a text" to contain a technical innovation (= invention) --
which seems very unlikely given its principle statements in the verdict --
this could then exactly correspond to the model of "software useright"

        http://www.freepatents.org/adapt/useright.pdf

proposed by J.P. Smets last year:  only a process is patentable, and if
licenses for executing it are charged, they must be charged at the user's
place, since the program as such is still not patentable while its
execution for a particular use is.

This useright model is actually quite consistent and feasible, far from a
mere formalism, as you can see by reading the article.

Still, it is a compromise proposal which is not in line with the EPC and
which would not be as good as keeping the rules of the EPC, as they are
explained by the BGH decisions from 1976-81 and recapitulated in the
recent BPatG decision.

Btw the BPatG also explicitely made it clear that its point of view is in
line with TRIPS.  They even have another special decision on that.  And
their latest decision explicitely refutes various recent theories from
both the EPO and the BGH, choosing to stick with older BGH theories, which
it considers more consistent with both the law and the practise of the man
skilled in the art.  The BPatG e.g. criticizes the "computer program as
such" and "computer program product" as phantom notions not familiar to
the man skilled in the art and not applicable.    

-phm