Re: [Patents] US patent reform strategy inaugurated by myths & lies

[PILCH Hartmut <phm@a2e.de>]

> >Personally I think that a better checking rule for prior art and a shorter
> >patent grant period (like 5 years) is something that can make a workable
> >solution.
>
> You raise the right issues.

As said before, both of these issues are irrelevant and just a detraction.

> IMHO the life span of a patent (20 years) is one of the main problem for
> softwares.

It can't be changed, and a look at the swpat database shows you that even
1 year for such patents would be a bad idea.

> I discussed a long time about this problem with Bronwyn Hall (Berkeley
> University), and it seems that american are finding too that a 20 year
> monopole create dangerous market imperfections. A right duration would be
> 3 to 5 years.

I can't see what good 3 years could do.
Certainly less harm than 20, but there is no market failure that calls for
even a 3 year swpat, as far as I can see.

> Unfortunately lawyers always refer to their Holy Text, the TRIPS treaty,
> saying that all patents last for 20 years.

Yes, and their whole current effort is about blurring borderlines, so that
even without the TRIPs treaty it would be impossible to reduce patent
duration for any specific subset of the patents.

> I think that patents aren't the right IP for softwares, a kind of
> utility model (short life span, non-exclusive licencing...) would be
> more appropriate.

Copyright and informal mechanisms etc taken together are quite
workable.

> Another reform to engage would be to charge exponential fees after the 3
> first years of the patent to dissuade companies to keep useless patents
> in their portfolio.

All this thinking is wasted, because the only reason why people are
pressing for patents is parasitism by vested interests such as those of
all the people sitting on SACEPO (standing advisory committee of the EPO).
There is no real need from any industry, not even embedded systems.

> But I agree that patents on hardware-related softwares (such as many
> embedded systems) could be granted in Europe since R&D costs aren't low.

The overall costs may not be low, but still patents are hardly an adequate
means to stimulate R&D in this area.

When we start talking about real hardware related research (not just
hardcopying software onto some medium where it is no longer copyable, but
actually researching into physical causalities), the economics do become
somewhat different.  But nobody has challenged the patentability of such
inventions anyway.

-- 
Hartmut Pilch, FFII & Eurolinux Alliance              tel. +49-89-12789608
Protecting Innovation against Patent Inflation       http://swpat.ffii.org/
120,000 signatures against software patents      http://www.noepatents.org/