[Patents] Gregs myth

[PILCH Hartmut phm@a2e.de]

> I disagree with Hartmut's disagreement with my argument that hardware
> and software are logically equivalent and therefore should be 
> equivalently patentable.

I did not disagree with your equivalence thesis.
I only disagreed with your conclusions.
 
> Hartmut and others should check out some of the literature on the
> latest generation of hardware/software codesign tools.  With such
> tools, you can describe some new digital system at a high level,
> and the tools will first partition the system into hardware and
> software aspects, and then using libraries of hardware and software
> components, generate specific hardware circuits and software source
> code to physically implement the system.  The tools can do this in
> multiple ways (for example, generating a completely software
> embodiment [by translating the system's VHDL representation into
> Ada], or a completely hardware embodiment].  To the tools, and the
> human user, the underlying hardware and software is irrelevant to
> understanding the functionality of the system being specified.
> A practicalality that the early computer science giants such as
> Turing, Post and Church would approve of.

That just means that these hardware patents should also not be awarded.
Anything that is not based on new insights into the working of the real
world but derived from known models should not be considered an invention.
 
> In that light, if the hardware coming out of these tools is
> patentable, so too should be the software.  

Also non sequitur.
First of all the hardware should probably not be patentable.
Secondly, it would still possible to make the hardware patentable for
reasons of economic expediency (which I think are not valid, but that's
arguable).

> It is not worth playing around with legal definitions to come up with
> a way to claim that software is still somehow different and worthy of
> a different sort of patent protection.  That is just bad science.

It is not science at all.
We are not investigating laws of nature but looking for useful rules.

But your method of reducing the question to "Are software and hardware
fundamentally different?" is bad philosophy.  Trying to reduce questions
of "what rules should we adopt" to questions of ontology is simply an
invalid way of thinking, typical for medieval scholastics.
 
> That many software programmers are upset about software patents I
> think has more to do with the extremely low level of quality of most
> of these patents issued by the patent offices.  The patent offices
> just don't have the expertise and resources to determine novelty and
> obviousness for patents.

I have never read a suggestion from you that would provide a means of
determining inventiveness (which is the European term for obviousness).
My suggestion is:  anything that can be derived using codesign tools or
other purely logical means is non-inventive.

You offer a lot of suggestions for improving novelty research, but
basically they all boil down to something like "Please hire ingenious
patentbusters like Gregory Aharonian".

Even if thousands of Gregs were available and the system was perfectioned
to its utmost extreme, we would still end up with a lot of uninventive
patents.  Even if the level of inventivity was raised, they would still be
economically harmful.

> If the patent offices did, I suspect the number of issued software
> patents would drop dramatically, and there would be far fewer 
> complaints about software patents, making much it easier to
> support that the logical equivalence of hardware and software should
> make them equally patentable.

That may be true.  But 
- it is quite unrealistic
- nothing follows from the logical equivalence
 
> So the question is: "Do we invest our time and energies trying to
> first create an artificial distinction between hardware and software
> and use that distinction to prohibit or greatly eliminate software
> patents", or "Do we invest our time and energies trying to get the
> patent offices to do a quality job with software patents".  I argue
> that the first question is both bad science and bad politics - the
> momentum is against the viewpoints of people such as Hartmut.

I argue that the second endeavour is futile and only serves to make people
like Greg Aharonian play a heroic role, while the first endeavor is
realistic, because it is supported by European law and by sound reasoning.
In Europe currently at least 40000 people share "viewpoints such as that
of Hartmut".

> It is inevitable that the EC will make there software patenting laws
> harmonious with the US.

There is nothing inevitable in this.  It is only the subjective will of a
lot of lawyers who currently happen to be swaying power.  And it is always
the tactic of those in power to attribute "inevitability" to their
personal will, thus lightening their burden of responsibility.

> That leaves the second question as the one to focus on, because it is
> good science (though admittedly not that much easier to achieve, given
> that large corporations in the US, Japan and Europe could care less
> about patent quality).

It is not only the large corporations who don't care.  The problem is also
- that there are not many highly motivated and skilled patent busters like 
  Greg in this world available for hiring by the patent offices, and the
  weakness of human nature as well as Parkinson's laws must simply be
  taken into account -- they are much more inevitable than those factors
  that you consider inevitable
- that even high examination quality wouldn't solve the basic problems.
   
> But as a scientist, I think it is false to argue that there is some
> conceptual difference between hardware and software to justify their
> being patent differently.  It is bad enough that lawyers engage in
> muddled thinking about patenting technology (here in the US, the Court
> of Appeals of the Federal Circuit which decides much patent policy has
> yet to fully understand the theory of algorithms and computing) - I
> would hate to see the scientists being equally muddled.

I hate to see scientists being philosophically unskilled.

-phm