gnu-misc-discuss
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: SFLC: "a penumbra"


From: Alexander Terekhov
Subject: Re: SFLC: "a penumbra"
Date: Mon, 18 Dec 2006 14:09:06 +0100

Alexander Terekhov wrote:
> 
> rjack wrote:
> [...]
> > One must be careful to define "Component[]" in context.
> 
> http://www.usdoj.gov/osg/briefs/2006/2pet/6invit/2005-1056.pet.ami.inv.html
> 
> "Although the court of appeals correctly held that software can be a
> component of a patented invention, it erred in holding that the creation
> of copies of software overseas, based on a master version provided from
> the United States, constitutes the supply of those software copies from
> the United States."
> 
> So I gather that accoording to USDOJ, supply by means of
> 
> http://www.research.ibm.com/quantuminfo/teleportation/
> 
> would NOT constitute "supply" for the purposes of 35 U.S.C. 271(f)
> either. Teleporters will undo 35 U.S.C. 271(f)! USDOJ should better
> inform the Congress know about that!!!
> 

Amazon proposes Chinese and French molecul test:

------
I. COMMON SENSE SHOWS THAT
A FOREIGN MACHINE PART IS STILL
FOREIGN EVEN IF ITS COMPUTER-READABLE
DESIGN IS FROM THE U.S.

As noted, if a machine is assembled abroad entirely
from parts made abroad, then no one would say that a
component of the machine was made in the U.S., even if one
of the parts was designed in the U.S. This is as true for
computers as for other machines. An optical disc made in
China from molecules supplied from China, is a Chinese
optical disc, even if its pits and lands are arranged in a
computer-readable pattern that encodes (stores) a software
program, CAD/CAM codes, song, or other information
supplied from the United States.

This “common sense” answer is illustrated with the
following two-part hypothetical assembly of a French key
and lock.

Part I: A French key has a unique pattern designed to
fit a matching pattern in a French lock’s mechanism. Both
the key and lock are made in France, entirely from materials
made in France. Not a single molecule of the key or lock is
traceable to the U.S. (See Figure 2).

No one would argue that this “key” component of the
key-lock assembly was supplied from the U.S., or that
Section 271(f) applies.

Part II: Now consider a new fact: the unique pattern
of the French key was supplied from the U.S. This pattern
(an example of engineering design information) was
conveyed from the U.S. in one of a variety of manners. For
example: (1) a U.S.-made master key is exported to France
where its unique pattern is decoded and duplicated
automatically by an electronic key duplication machine to
make the French key (see Figure 3), or (2) CAD/CAM
computer codes are e-mailed from the U.S. to France where
they are used to program a machine to manufacture the key
to the unique design specified by the U.S. engineer. No
matter how the U.S. pattern is supplied, all of the molecules
(matter) of the replicated French key are still supplied
entirely from France. Only the design information was
supplied from the U.S. and since information is not a
physical object, Section 271(f) plainly does not apply.

As this hypothetical illustrates, the above-proposed
“Molecule Test” provides a bright line test for anyone
concerned about possible liability under Section 271(f): if
the foreign assembly does not include a single molecule
exported from the U.S. by the potential defendant, then
Section 271(f) does not apply. There rarely, if ever, will be
uncertainty on this point.

[...]
Second, on the question of what Microsoft
contributed to the foreign computer assemblies, the district
court used the terms “code” and “software” to refer at times
to information and at other times to matter. For example, the
district court noted “the undisputed fact that the object code
is originally manufactured in the United States” AT&T, 2004
WL 406640, at *7. Its use of the term “manufactured”
suggests that the district court had in mind physical discs, as
products, not information, are “manufactured.” But its
reference to the “object code” elsewhere may be directed to
software information (e.g., a sequence of binary numbers),
see AT&T, 2004 WL 406640, at *4 (“software or object code
contained on the golden master disks”).

The Federal Circuit made the same mistake. It failed
to carefully distinguish between things and the design of
those things. Thus, it mistakenly analogized software
information to liquids and gases. AT&T Corp., 414 F.3d at
1370-71. Software information is not akin to liquids and
gases because it has no mass and no molecules. Its
information content is transferred from disc to disc without a
single molecule being transferred—just as the information in
this Brief is transferred to a photocopy without a single
molecule being transferred.
------

regards,
alexander.

--
"Please do not buy from Amazon"

           -- Richard Stallman


reply via email to

[Prev in Thread] Current Thread [Next in Thread]