Re: [Patents] U.S. Office Joins an Effort to Improve Software Patents

[xdrudis <xdrudis@tinet.org>]

El Fri, Jan 13, 2006 at 02:40:12AM +0100, Stefano Maffulli deia:
> http://www.nytimes.com/2006/01/10/technology/10blue.html?emc=eta1
> 
> Interesting reading, especially since the reporter identifies Open
> Source with IBM :)  I would add that this is one of the reasons FSFs
> insist on using the term 'Free Software' and its translations.
> 
> Going into the content of the article, what do you think of such
> efforts?  Pure waste of time and money or can this be a piece of an
> answer to the bad-patent proliferation?  
> 

I've read the story elsewhere, I don't have internet here now to read 
nytimes. 

I think it's impossible to fix the patent system while pretending that
any subject matter is patentable and then trying to fix the rest of
criteria.  This is an effort to improve the tests of novelty (and
maybe obviousness?).  If they had done something similar for other
fields of endeavour, the patentable fields, it might have helped in
those fields, but it this case it will just waste efforts and possibly
streghthen software patents.

Currently I think Aharonian said about 50% of software patents are
ruled invalid. This means the threat any random software poses is
considerable but still relative. If they managed to improve the
quality of software patents it would just make the threat more
certain. It might decrease their number or not but it would increase
their severity. A centralised prior art database has been discussed a
few times, and I agree with those thinking that it would mostly help
patentees to produce more waterproof patents, not less of them.

It would place a burden on programmers to defend from an unjustified
attack by putting their weapons on the table for their attackers to 
better plan their strikes. It might even induce someone to put more 
doubt on nonregisterd prior art. 

Even if they were fewer, and even if they would by some magic become
all new, non-obvious, and as brilliant as you could dream, they would
still impede innovation instead of promote it, they would still
fortify monopolies for no good reason, they would still prevent
freedom of expression, of commerce and of beign informed. They would
just be harder to fight in court.

They say the USPTO has ruled the MS FAT patent valid. Well I won't
comment on whether it is new or obvious or not, but if it isn't so
what ?. Everything has possibly been new and non-obvious at some point
in time. If it had been patented then, would it cause less disturbace
than it does now ? The problem is that the effort you need to innovate
in software is already worthwhile enough without patents, that
software is to "soft", too ubiquitous and to useful to reuse and
recombine to be subject to broad monopolies like patents, that
software is information and the patent system intent is to foster
spread of information, not restricting its diffussion, that software
is speach, etc. You know all that already.

If OSDL, IBM et al can't see it, then maybe the only good result of
this exercise would be to bring some more quality to the
newness/obviousness tests (however unlikely) and then leave still in
place the basic problems of swpats, but maybe in that hypothetic
situation those patent panteists will see the problem is not in
obviousness/newness. This is wishful thinking, though.

Having said that, if the system allowed some form of timestamping
prior art which would cheaply but strongly prove it had been published
at some time, while at the same time not building a searchable 
repository of prior art, it could be of some use. But this is not 
what they plan from what little I've read.

It's like if you go to your neighbour to complain it's playing his
music too loud and you can't sleep and he offers you to bring some CD
of music you like more for him toplay instead of his CDs. Nice, but
you still have a right to some silence (and to keep your CDs), and the
point wasn't musical preference but music volume.

-- 
- jo també vull una Europa lliure de patents de programari  -
--------------------------------------------------------------
 EuropeSwPatentFree - http://EuropeSwPatentFree.hispalinux.es
--------------------------------------------------------------
http://patents.caliu.info

Xavi Drudis Ferran
xdrudis@tinet.org