Re: [Patents] newsscan: Software patent limits 'go too far'

[Bernard Lang <Bernard.Lang@inria.fr>]

On Thu, Oct 23, 2003 at 10:56:33AM -0400, Russell McOrmond wrote:
> 
> On Wed, 22 Oct 2003, Bernard Lang wrote:
> 
> > The point is : there are not equivalent
> 
>   Please detail from a Legal or Technological point of view how they are
> not equivalent.  How does it matter if programming is interpreted by a CPU
> at run-time or whether programming is burned into an FPGA?

legal is not the issue ... legalities are arbitrary

technological ...  you know the difference ...

but the point is economics... and the reason for patents is also economics.

>   I also disagree that the economics are different, but that is a
> conversation we don't need to get into as the economic situation of each
> isn't really all that relevant.

The economics of machines is rival ... no way it can be disseminated
wiothout cost for each copy ... hence the need for production and
commercial activity  .... which may justify patenting to recover costs.

The economics of software are non rival.  It can develop without
commercial activities.

>  Some engineers (and/or their employers)
> say that the economics are different, but then again some engineers in the
> "software manufacturing"  lobby say that the economics is the same (and
> that any methodology other than "software manufacturing" such as FLOSS
> doesn't exist, that the earth is flat, that...), so why bother focusing
> only on these arguments?

because I try to stick to reason ...  even when others will resort to
lies or religion (or ideology).
  On any other ground, anything is true and false, and then only big
sticks matter.

> Note: I am not at all suggesting that the economics of hardware and
> software are the same.  What I am suggesting is that the programming of an
> FPGA is software, not hardware.

I have not spent enough time tinking about FPGA (any reference
around), but I tend to agree with you ... until further thoughts.

I did think about a similar problem, though, and that is printable
circuits, i.e. circuit (integrated or not) that can be produced very
cheaply with a conventional printers from supermarkets.  To me the
economy of that is essentially the same as that of software, and it
should be excluded from patentability.  ...  One way is to say that
private production of products for personnal use is not subject to
patents ...  and the logicial description is not patentable.

>  As more and more hardware becomes 
> programmed with software then things which people think of as hardware 
> must be treated more like software (and not the other way around -- 
> software should not be treated like hardware).

I agree ... but this has to be formulated with clear definitions, so
as to come up with economically sound answers.
  I think it is unwise to confuse the issues ... we must clearly
distinguish software and hardware, as was done in the amended
directive.  The we must have clear rules so that we get the most
effective patenting system from an economic point of view
... effective meaning that it will best favor innovation, economy and
welfare.

>   When there are no longer innovative processes by which physical hardware
> is built because all of the innovation moves to programming, then there
> should be far fewer patents granted as only the processes used to produce
> physical hardware should be patentable, not any programming of
> pre-produced hardware.  Two differently programmed FPGA's are the same
> hardware that were manufactured using a single (possibly patented)  
> process.

> > The role and justification of patents is economic.
> 
>   The economic justification used by the United States (to further the
> arts and sciences) is just one thread seen in the world.  There is also
> the rights-based justification:
> 
>   Article 27(2) of the United Nations Universal Declaration of Human 
>   Rights says:
> 
>     Everyone has the right to the protection of the moral and material 
>     interests resulting from any scientific, literary or artistic
>     production of which he is the author.

yep ... and this is nonsense where granting patents is concerned.
When 2 people invent the same thing, which happens often enough, there
is an arbitrary rule to decide who gets it, which is not the same on
both sides of the Atlantic.  How can this be considered a human right
when it is arbitrary ?
  Furthermore, we all know that some of these productions do not entail
any material rights ... such as the production of new scientific
understandings.
  Finally, where do software manufacturer list the names of software
authors, who do have moral rights on the software according to this
text.  And I would emphasize that this text is about human rights, not
about corporate rights.

   So the only sensible way to interpret this text is to say that,
once such "material interests" have been granted, for whatever reason,
they should be protected for all.  But it does not say much about the
granting.

  And I am afraid the text does not say much more about moral interets.

  This is a bad text (in the sense of poorly written) because some
people tried to overextend it, thereby emptying it of any sensible
meaning.  A tradition of the IP lobby, which was probably at work there,
while we were not paying attention.

>   I for one am not debating whether an inventor should receive the moral 
> and material interests resulting from an invented process.  What I am 
> debating is what constitutes moral and material interests, and what 
> constitutes an invention.
> 
>   Two otherwise identical computers programmed differently, or two
> otherwise identical FPGA's programmed differently, can not represent
> different invented processes.  Information processes or other
> non-manufacturing processes do not represent inventions, a private citizen
> (outside of a manufacturing process)  creating the same outcome should not
> be considered in violation of a patent, and so-on.

it seems that we agree ... the question is to structure this into a
consistent set of simple principles ... that do not disagree with
existing texts.

> > Patentability is there if it encourages investment is innovation,
> > disclosure and industrialisation. The economics of hardware and software
> > are different, even though functionality may be the same.
> 
>   Patent laws are acts of parliament that are interpreted by judges, not
> economists.  The purpose of law is to create stability and certainty: you
> should know when you are breaking a law and when you are not.  Economics
> is anything but stable or certain.
> 
>   While economics plays a very important roll in the justification for
> patent law, and is critically important for creating such laws, it can not
> play a useful roll in the interpretation of existing law.

mostly true ... within the limit of what jurisprudence can do, basing
decision on whatever it deems appropriate.

so what ?

> > Would you ask turkeys whether thanksgiving is a good idea, or
> > shoemakers whether walking barefoot is a good idea.  Don't ask Greg?
> 
>   It is a given that there are vested interests in this discussion, and
> not just from one side.

Certainly (though I am losing money, if anything, in all this) ... but
that is not the point.

While coporations on one side, and sme/technicians/public on the other
side have vested interest in what the law says.  They are all
competing economic actors in this exchange of goods and resources, as
producers and users.
  IP lawyers are parasites to the process.  They do not provide
anything, but want to change the rules to get more blood flowing in
their direction.

   well actually, IP lawyers do play a role in implementing
regulation, once consensus is established.  They can even provide
input regarding the design of rules so that they can be well
implemented.  They have no competence (either technically or
constitutionally) to orient the choices one way or the other.


> We need to move beyond that to determining what
> makes good public policy and how to enact laws that will help create
> certainty.  Even at this stage with the recent developments the legality
> of software patents is uncertain in Europe, and thus the law is still 
> flawed.

Interestingly enough, those who made the most use of this argument
produced a directive that was intentionally unclear and lacking proper
definitions.

>   It should be simple to state that while the existence of software within
> an otherwise patentable manufacturing process does not make the process
> unpatentable, that software alone can not represent a patentable process.

precisely what the european parliament did

>   There are many ways to try to say this including the "as such" and
> "technical requirement".  Until patent offices and patent courts agree
> with other patent policy makers that software alone should not be
> patentable then software alone will remain patentable.

point is, courts seems to think (in Europe) that software alone is not
patentable.  That is precisely why the patent lobby wants to change
the law, since courts have the last say on this.

>   Complaining that by our interpretation software should not be patentable
> in Europe is not helpful given that European patent offices and courts
> seem to disagree and it is their interpretation (and not ours) that
> matters.  We need to continue to push to enact laws until such point as
> patent offices and patent courts interpret the law as intended.

yes


Bernard 

-- 
         Non aux Brevets Logiciels  -  No to Software Patents
           SIGNEZ    http://petition.eurolinux.org/    SIGN

Bernard.Lang@inria.fr             ,_  /\o    \o/    Tel  +33 1 3963 5644
http://pauillac.inria.fr/~lang/  ^^^^^^^^^^^^^^^^^  Fax  +33 1 3963 5469
            INRIA / B.P. 105 / 78153 Le Chesnay CEDEX / France
         Je n'exprime que mon opinion - I express only my opinion
                 CAGED BEHIND WINDOWS or FREE WITH LINUX