Re: [Patents] Software patents: the case against

[Erik Josefsson <ehj@cleosan.com>]

Christian Beauprez wrote:
> Hi everyone,
>  
> I have written a report on this issue that is available in both .pdf and 
> .html form. I know that many of you will be familiar with the arguments 
> already but I have tried to keep it as concise as possible.
>  
> it is available at: 
>  
> http://www.beauprez.net/softpat/legalfacts3.html
> http://www.beauprez.net/softpat/legalfacts3.html

The Petterson queue patent comes to my mind. It's about a queue ticket 
apparatus, like when you buy a ticket in a railway station.

The application was rejected by the Swedish Patent Office, the Swedish 
Court of Patent Appeals and the Swedish Supreme Administrative Court. 
Their judgment was: it is an algorithm that solves the problem to 
organize a queue in a certain way. An algorithm is not an invention. The 
achievement has no technical character whatsoever.

Later the EPO granted a patent with the argument:

"The subject-matter of Claim 1 represents a hardware-like selfconsistent 
physical structure, which has a concrete technical construction... Such 
a technical realisation of the claimed system forming subject-matter of 
the claim is by no means a mental rule or method which is only "as such" 
excluded from patentability by Articles 52(2) and (3) EPC."

This sounds really serious, but when you have a look at the drawings of 
the patent you just can't believe your eyes. The story is also that the 
patentee filed further more detailed drawings and descriptions, as if...
... well, as if the solution became more technical the more physical 
details were added.

So in EPO 1994, the evacuation of meaning of 52.2 exclusions as 
achievements, replacing achievementws with abstractions, was already 
under way.

Such a dramatic change of interpretation should be brought before the 
EPO EBA, and the clash between Swedish and EPO interpretation was a 
great opportunity to raise the question. This was never done:

http://www.sslug.dk/emailarkiv/patentdirektiv/2002_05/msg00038.html

It is still a repeated argument, that each swpat case is individual, and 
no generalised question from swpatenting can be brought to the EBA: The 
question of what is technical *must* be left to the TBAs to decide.

And only the TBAs can decide what the legislator really ment with "as such".

I am not sure if extra judical concepts are ever introduced in law, but 
clearly, legislators should rather try to involve the concept of the 
Turing Machine into law than "normal interaction between computer and 
computer program". Normally, turning the computer on is not very 
difficult, but it may be non-obvoius.

But then again, did not Ericsson have a patent on turning a machine on? 
Or was it Nokia on saying Yes? Some inventions are indeed brilliantly 
trivial.

//Erik

here is a shorthand of 52.2:
http://elis.ugent.be/~jmaebe/swpat/cii.html#broken