From phm@a2e.de  Mon Apr 30 19:14:01 2001
From: phm@a2e.de (PILCH Hartmut)
Date: Mon, 30 Apr 2001 20:14:01 +0200 (CEST)
Subject: [Patents] Re:  Patents on user interfaces.
In-Reply-To: <Pine.LNX.4.33.0104301902140.2285-100000@mulix.oas.a2e.de>
Message-ID: <Pine.LNX.4.33.0104301959320.2541-100000@mulix.oas.a2e.de>

I answered to an answer of Greg to Ramos.  My impression had been that
Ramos wanted new people from the EPO mailing list to answer the question.
But instead Greg jumped in, and I was quick in my usual refutal of Greg's
arguments (already exercised about 20 times on patents@aful.org during
the last year).  Unfortunately too quick, leading to some bad typos which
I have to correct here:

> > >I would like to know your opinions about whether user intefaces should
> > >be patentable or not.
> >
> > User interfaces are one aspect of the field of software technology.  In
> > all cases they are functional and utilitarian.  In all uses, they
> > achieve a technical effect (once once formally defines technical effect).
>
> .. once you define technical effect as "functional and utilitarian" or as
> "practical and repeatable solution" as seems to be the consensus of the
> EPO's current Technical Boards of Appeal, see
>
> 	http://swpat.ffii.org/vreji/papri/jwip-schar98/
>
> However this interpretation of "technical" has its drawbacks.  It makes
> almost anything patentable, and it makes Art 52 of the EPC completely
> meaningless.  It is, although in itself quite consistent, in contradiction
> to current society's basic consensus on economic policy (e.g. it makes
> business methods patentable, although almost all politicians reject that
> idea) and to the law.
>
> The old European doctrine that explains technical as "belonging to the
> field of applied natural science" or "teaching new insights about the
> causal relations between natural forces and a physical effect" does not
> have those drawbacks.
>
> > User interfaces are equally implementable as either hardware or software,
> > that is, one can write VHDL to specify the UI and then have the VHDL
> > automatically converted to either hardware or software, or a mix of the
> > two.
> >
> > If you allow someone to get a patent on a digital hardware implementation
> > of the UI (which could be sold as a co-processor like other types of
> > application co-processors being sold), digital hardware which should take
> > about as much time and effort to design as a software version of the same
> > UI, then to be equitable under the law you have to allow a patent on a
> > digital software implementation of the same UI.
>
> You could say it the other way around:  if the ideal embodiment of a
> patented invention according to the current state of the arts is something
> immaterial, e.g. software or business method or other type of logic,
> then you should, for consistency's sake, not grant the patent, even if a
> hardware implementation is also possible.

On the other hand, I may add, consistency is not as high a priority as
some people selectively pretend.  If it was found that it was good for
society to grant patents on hardware implementations of logical ideas but
not on the software versions, then why not.  There could even be cases
where one might be tempted to thinkg that.  E.g. Fraunhofer/Thomson's
charge of a few cents of license fees on each MP3 hardware device is not a
big economic problem, while the impossibility of developping opensource
software for MP3 encoding (you have to pay a lumpsum of 1 million USD for
a non-proprietary development right -- this is economically paradox)  may
be depriving society of a lot of interesting options.

> This rule is equivalent to the traditional "natural forces"
> doctrine mentioned above.

> And if you allow immaterial embodiments, you enter the realm of
> information economics.  Information economics is different from
> market economics.

> The patent system is based on a duality of free information
> (public disclosure) and monopolised material emobidement.  As soon as this
> duality stops to work, you run into trouble.  This is not to say that the
> traditional patent system, which was limited to empirical
> (technical) knowledge and excluded abstract-logical ideas, did not
> have any trouble.
>
> > On the other hand, if you are against patents both for digital hardware
> > and software, then one can argue that user interfaces are not worry of
> > .... whoops ..... are not worthy of software patent protection.  But
> > then one would pretty much have to argue against the patentability of
> > analog hardware as well (can you look at a circuit board and tell
> > which ICs are digital and which are analog?).  Which means that you are
> > pretty much against all electronics patents.
>
> It may well be that if you disallow logical ideas (the realm of pure
> reason, anything that calls for mathematically proof rather than for
> experimental verification) you will end up abolihing many electronics
> patents.  But not all.  I have done a test on a sample of 100 recent
> EPO patents, and I find that applying the logical/physical (matter
> vs mind, empirical knowledge about natural forces vs pure reason)
> boundary will lead to the rejection of only about 3% of the currently
> granted EPO patents on grounds of lack of technicity.

> I believe that normal patent lawyers should support our cause and help
> us to argue for this boundary.  It harms the patent system if you
> allow it to be taken into troubled waters by a few specialists who
> make their living on software patents.  All those who go along with the
> "anything useful" or "all practical and repeatable solutions" doctrines
> of the current USPTO and EPO are really helping to remobilise an
> anti-patent movement that has always existed as an under-current and
> that has had most well-known economists on its side (including people
> like Arnold Plant, Friedrich Hayek and Fritz Machlup).  Any victory
> of the patent lobby over the wishes of the software community will
> be a Pyrrhean victory.

-phm






