Most recent full copy of Directive on patentability of computer implemented inventions

[Niall Douglas]

On 15 Aug 2003 at 19:26, MJ Ray wrote:

> Being defeatist is the single best way to ensure that we get a form of
> software patents.  Devil and the deep blue sea.  Given this, I'm
> arguing that "These are crap because X, Y, Z but please at least don't
> let them do A, B, C."

I think we already have software patents. The directive claims to 
clarify the situation so it too acknowledges that they already are in 
effect here. It's just that no one can really say /how/ they are in 
effect yet which is what the directive is supposed to sort out.

> Core is that I regard programs as a branch of mathematics, discoveries
> not inventive acts.

Programs aren't mathematics because they are eternally useful ie; do 
something without the involvement of a human. Therefore they should 
reside in a totally unique category from either mathematics or 
anything else. DNA sequences however would fit right where computer 
programs go.

> I also look forward to patenting electronic documents, etc, if they
> really screw it up.

It's why we need the "forum to register abuses" amendment. Then you 
could patent your document, and register a complaint against yourself 
for doing so :)

> And yet, we are often told that one reason for this is to reduce the
> ambiguity in the current system.

Too many cooks spoil the broth.

> [...]
> > The third highly important amendment is needing to set what
> > precisely involves an inventive step. I would make it high ie; "a
> > substantial advance over the status quo".
> 
> Is this not fixed already?  I thought you said you couldn't change EPO
> rules.  Immovability of this has been suggested as one reason why we
> must have them voted down.

EPO rules on paper are quite high. In practice, those rules appear to 
be almost unenforced. This is probably why most MEP's are mistaken in 
thinking that the EPO is a responsible patent office.

I don't see why an amendment couldn't mandate all EPO software 
patents to have to run through a board of say ten independently 
selected experts in software. Problem would be, who pays?

This can be argued in that the hundreds of millions of euro saved by 
SME's in legal costs in not having to take software patent 
infringements to court are well worth paying a few hundred thousand 
in board fees annually.

Cheers,
Niall