Re: [Patents] Patenting computer-using inventions under the EU Software Patent Directive

["Christian Beauprez" <beauprez@beauprez.fsnet.co.uk>]

Dear Holger,


> Siemens, Philips and others in the 20% non-SME segment of the
> European software industry repeatedly have voiced there concerns
> about the directive (in its Parliament version) excluding too much
> of their current portfolio. Examples I have heard repeatedly were
> claims on computer tomography, anti-blocking systems, GSM, and so on (and
> naively seen in an ideal world making complex hardware-heavy systems
> is the market that big industry should be good at).

It seems to me if they want to patent an improved anti-blocking system that 
applies no physical effects and the innovation lay solely in data processing 
that would indeed be unpatentable by the Parliament's directive. But then 
the question that has to be asked and which the patent offices are no longer 
asking is "does this claim relate to a computer program " . I like to take 
the traditional view on this that if the invention lies only in instructions 
given to a data processing system, with no significant post solution 
activity, the claim must fail .

To quote Justice Laddie,
"Mr Birss is right that a computer set up according to the teaching in the 
patent application provides a new "tool" for modelling crystal structure 
combinations which avoids labour and error. But those are just the sort of 
advantages that are obtained by the use of a computer program."

I think the problem happened when we had the terminology "as a whole" 
introduced in swpat law. After this point the substance of the claim was 
ignored and the question just became "is it new, non obvious and useful?" 
(the victory of "technical" over "statutory subject matter" or "invention".) 
People have to be reminded that even if something is all of those things 
(new, non obvious ,and useful) it still doesn't make it an "invention" in 
the sense of patent law.

> Now, most of these systems are heavily hardware-dependent and
> I have been asked whether we have a list of showcase patents (preferably
> of inventions involving a computer somewhere, thus being CII/SEI)
> that are *not* affected by the exclusions of data processing, the
> requirement of forces of nature etc in the Parliament's version.


I know one : ABS comes to mind.

> A non-academic outcome of this exercise would be: see, the claims
> xx,xx,xx of invention yy,yy,yy are not necessarily excluded by
> the Parliament directive, what do you clamor about? (ignoring of
> course the question of novelty and inventive step which could be
> taken for granted to simplify the discussion).

I agree the question is what is statutory, not whether it is novel or 
unobvious.

> If sb already has such a list for "separating mechanics from
> data processing" for defensive purposes, that would be a good starting 
> point,
> if not, who would like to (join) compiling it (below is a starting
> point by J Bosson)? (The case might be made that's it's too late
> for theoretical exercises at the moment of course.)

You know, I actually think that despite minor philosophical tustles within 
the movement against software patents there are core things that everyone 
can agree on:

The parliament directive is a document that clarifies the law and obtains a 
good balance overall.

The commission document is nothing more than an enabling act to introduce 
unfettered software patentability.

We want people to understand this and vote accordingly in the interest of 
clarity, harmony and the status quo of Article 52 EPC (not the patent office 
dogma), and to insure that the internet and software remain a "free press" 
for business and the public.

I think there is also an argument that instead of "why don't we apply 
patent law to x if it does the same as y" it should be "which system in this 
type of x  innovation  will work better". This is the debate that they can't 
win.

Sincerely,

Christian