On 5/20/07, [EMAIL PROTECTED] <[EMAIL PROTECTED]> wrote:
A process is a flow of materials thru a set of machines. (It may
actually be more general than that, but let's just consider this
example for the moment.)
This flow of materials can involve decision points. But basically
it connects together discrete operations.
In software we can represent materials with symbols. And the
symbols can be made to flow in exactly the same way as the
process above. These are operated upon by discrete operations.
So how is this different? Simply because we have replaced screws
with electrons? It looks like an algorithm to me.
Yes, it is. But with the line we drew, the infringement of a patented
process lies in the physical execution of the algorithm, not in the
the metaphysical algorithms used to implement it. Maybe try it this
way: The U.S. Patent & Trademark Office has published algorithms
(processes established by regulation) regulating the flow of patent
applications through their system. Should its algorithms be patentable
subject matter (ignoring factors such as prior art and the heighth of
the innovative leap)? I.e., should one be able to patent the creation
of patents?
This isn't a far-fetched hypothetical example. Since the Federal
Circuit approved of software business process patents, there has been
a tidal wave of business process patents issued including a couple
that allowed patents on discrete laws. Isn't the reductio ad absurdum
the patenting of methods of creating algorithms, i.e., patenting
mathematics? According to the Federal Circuit, the deciding factor is
whether the algorithm is "useful" or "abstract." If it is "useful," it
can be patented. So binary mathematics is now patentable subject
matter and is routinely being patented.
Our response is that if it's information in/information out, it's
non-patentable. You can patent the physical steps and their sequence
in a process but not the algorithms themselves. The European treaty on
patents might illustrate this principle better. I'll omit irrelevant
portions without indicating the snips:
(1) European patents shall be granted for any inventions which are
susceptible of industrial application, which are new and which involve
an inventive step.
(2) The following in particular shall not be regarded as inventions
within the meaning of paragraph 1:
(c) programs for computers;
(d) presentations of information.
(3) The provisions of paragraph 2 shall exclude patentability of the
subject-matter or activities referred to in that provision only to the
extent to which a European patent application or European patent
relates to such subject-matter or activities **as such.**
<<<
First, note that the language is addressed specifically to computer
programs, not to whether they are implemented in a ROM or in software.
I can't tell you how many drafts we wasted attempting to address the
problem of programs implemented in hardware rather than in software
before we caught the language that drew no distinction between
programs implemented in hardware or in software. I'm sure it was more
than 20 drafts. And that realization guided us to what we believe is
the correct bright line betwixt the physical and the metaphysical,
between patents and copyrights. E.g., you can patent the physical
means of putting marks on a sheet of paper, but you can't patent the
sequence of symbols themselves; that is the province of copyright law.
The original intent of the treaty and its interpretation by the the
European Patent Office was that patentable processes could be
controlled by computer programs but that the programs themselves could
not be patented. The European Patent Office, which derives its
revenues from patent application processing fees later reinterpreted
the treaty to allow precisely that which is forbidden by it. Our task
was to draft language that resolved the ambiguity introduced by the
treaty's later misinterpretation by those with a vested interest in
ignoring the treaty's restriction.
The language used in the treaty may seem convoluted, but there is a
very practical reason for it. The major patent systems of the world
all require that nothing can be described in the "claims" portion of a
patent application, since that is intended as the wording of the final
patent itself. The treaty language addresses that problem by, in
effect, saying you can include a description of a computer program's
functionality in the claims to the extent that it is necessary to
understand the physical steps of a process, but that doesn't give you
a patent on the computer program. So while the language might seem
unnecessarily convoluted to those who have not studied patent law, a
patent lawyer will (or at least should) understand instantly why the
treaty language is worded the way it is.
And looking at it from the other side, if I patent the "process",
does my patent then prevent competitors from running software
which precisely mirrors the operations of my process?
No, but that does not allow competitors to mirror the physical
operations of the process. You have to look to copyright law to
prevent them from running the software. Note, however, that copyright
law allows a lot of flexibility in this area. E.g., you don't get far
in copyright law with arguments that a copyright encompasses methods
and concepts. That is to say, you can copyright your publication of a
recipe for roast duck, but you can't stop people from using your
recipe to roast their own ducks or from publishing their own recipes
that are derivative works. The reason is that you can copyright your
description of your recipe to the extent it is original work and
non-factual, but you gain no monopoly on the methods and concepts
described. We watched that issue being briefed in the SCO v. IBM
litigation, which led to SCO abandoning its copyright claims.
Is it even possible to define the size of an innovative leap?
It is so incredibly subjective. Everybody will have a different
view of each innovation.
I don't think so, like I said the courts are now into the third
century of trying to do so in the U.S.
But I digress. Bottom line: I don't know how to judge innovation.
Not my own. And often not others'.
Exactly.
I am. Altho some of the terms you use strain my brain and leave
me unsure just what you are saying.
+ I think I know what "information" is. But I'm not sure about
"information state".
I meant the physical state of a physical device or carrier wave. For
example, the physical state of a gate in a transistor, the binary
information state of the machine as opposed to the information thus
stored. We drew our line betwixt the two. To understand it you have to
study the interplay of the inclusions and exclusions in the individual
subdefinitions in the Groklaw Definition.
<http://www.groklaw.net/article.php?story=20050531195311724&query=groklaw+definition>.
+ I know what I mean when I say "carrier wave". What do you mean?
They don't seem to be the same.
Ah, I did leave out that detail, didn't I? Sorry. Wikipedia defines it
as a "waveform suitable for modulation by an information-bearing
signal." We eliminated the "wave" from "carrier wave" from our
definition because of things like the "entanglement" phenomenon in
quantum mechanics.
<http://en.wikipedia.org/wiki/Spooky_action#Background> ("Einstein
famously derided entanglement as "spukhafte Fernwirkung" or "spooky
action at a distance").
I.e., will wave theory even be involved in the relevant aspects of the
quantum computing future? It seems as though at least in some aspects
wave theory may not apply. See
<http://en.wikipedia.org/wiki/Quantum_teleportation> ("Quantum
teleportation does not transport energy or matter, nor does it allow
communication of information at superluminal speed, but is useful to
quantum communication and computation"). I am not very far into
quantum mechanics, but my sense from the little I know is that wave
theory's involvement may be either less applicable or irrelevant in
the quantum computing future many are working toward.
So we developed this sub-definition:
"'Carrier' means a force of nature manipulated for purposes of
conveying or storing information, but shall exclude any information so
conveyed or stored."
We thus sought to sidestep the issue of whether wave theory even
applies in quantum teleportation.
<http://en.wikipedia.org/wiki/Quantum_teleportation> ("Quantum
teleportation does not transport energy or matter, nor does it allow
communication of information at superluminal speed, but is useful to
quantum communication and computation"). We sought a definition that
would apply to all known and foreseeable means of automated data
processing. E.g., mechanical computing, hydraulic computing,
electronic computing, optical computing, quantum computing. And we
wanted the line to be valid even at the device level, e.g., we now
commonly store data electronically in transistors, magnetically, and
via pits on disks read by optical devices. But we were also aiming for
a definition that was future-proof.
"Force of nature" is a term that is already well defined by patent
case decisions, so we used it.
The only bug brought forward for which we felt that we might not have
a good answer was the comment that argued that in quantum mechanics,
the entire universe is information, not energy or matter. I think that
was a comment directed to the entanglement issue, but we didn't have a
quantum physicist on the team so didn't have a good response. So I
suggested that the courts would probably have no difficulty working
within the confines of the broader view that there are such things as
energy and matter that are distinct from information. If there is a
quantum phsyisicist lurking among us, I'd love to acquire some
enlightment on that issue.
Best regards,
Marbux
_______________________________________________
EUGLUG mailing list
[email protected]
http://www.euglug.org/mailman/listinfo/euglug
