[Patents] logically provable vs experimentally verifiable

[PILCH Hartmut phm@a2e.de]

This didn't reach the mailing list for days due to some error in the name
service at aful.org.  Although it badly lacks novelty by now, I'd still
like it to get through to the mailing list archive.

---------- Forwarded message ----------
From: PILCH Hartmut <phm@a2e.de>
Subject: Re: [Patents] (no subject)
To: Gregory Aharonian <srctran@world.std.com>
Cc: patents@liberte.aful.org, Ram'on Garc'ia Fern'andez 
<ramon@jl1.quim.ucm.es>

> >These "hardware" patents do not need any physical experimentation to
> >be verified.  They belong to the logical world.  They can be deduced
> >with paper and pencil only.
>
> Which is true for large numbers of chemical and mechanical and circuit
> patents as well, which can be designed on a computer as easily as
> software.

Ramon said "verified".  Not "designed".
What do you mean by "as well" ?
Maybe the ones you mean are exactly those that he means.

> So once again you propose a criteria that shows that
> software is no different from other areas of technology.

No.
The proposed criteria eliminate about 3% of all European patents,
as we found out by applying them to random lists of EPO patents.
There is no such thing as a difficulty of distinction.
The borderline is very clear.
I didn't know that before, but after investigation I know it now.
We can deepen the investigation together, if you like.

> That is why the European governments are pretty much conceding the
> patentability of software - they realize that you can't create
> unambiguous laws to differentiate software from all other technologies.

They are quite clueless on the issue, since they are occupied with the
art of politics, which is quite different from the art of studying these
questions.

This however does not mean that european governments have believed your
bragging about co-design-tool cosmology.

Politicians and lawyers may be quite foolish, but they do know that
borderlines in real life are hardly ever clear, and that some lack of
clarity in a borderline doesn't mean that one can renounce to borderlines.

They know for example that the borderline between killing, theft, lying
and bragging can be quite fuzzy, but they wouldn't use that as a reason to
hand down a "uniform sentence to everybody".

> Designability criteria like the above fail, economic arguments fail
> (other areas of patenting, like drugs, have far more serious economic
> problems due to patents), incentive arguments fail, and logic
> arguments fail (software is technical in any form).

None of these fails.  But when someone is selectively picky on a certain
issue, he can of course reject any reasoning.  Reasoning depends on a
certain mutual consensus of wanting to find the truth.  It is normal that
some people do not subscribe to this consensus.

> The courts in the US struggled for years and eventually realized the
> impossibility of practically discriminating against software.

Upholding the distinction requires some intelligence.  In the US patent
system, there was not a great incentive for people to come up with this
intelligence.  Patent lawyers didn't have much reason to try to uphold the
distinction.  There was a massive interest on the part of the patent
movement, including people like Greg Aharonian who owe their living to
crappy patents, not to discriminate.  More than that, there was even an
economic policy called "pro-patent", with a goal of using the patent
system to conduct economic warfare against countries like Japan by
erecting a kind of informal trade barrier.  (I am just reading a Japanese
book on this economic war.)

> It is why I still think the argument here is not against software
> patents, but against all patents.  Because if you can't distinguish,
> that's the remaining issue.

You are constantly trying to reduce the question to "either all or none",
hoping that politicians will choose "all" because they don't dare to say
"none" (or can't, due to TRIPS).

This is a very primitive strategy, and I don't think anyone except those
who are already deeply entrenched in the patent movement, can be deceived.
So why reiterate it all the time?

Our studies have shown you wrong.  The borderline is clear even though
there is no necessity for it to be clear.

-phm