Re: LZW patented file left in .orig.tar source package?
On Thu, 2002-10-24 at 00:36, Jeff Licquia wrote:
> On Wed, 2002-10-23 at 19:34, David Turner wrote:
> > I found a case which says that blueprints are components in the sense
> > meant by (c) (well, actually (f), but it's the same language) above:
> > Moore U.S.A. Inc. v. Standard Register, No. 98-CV-485C(F), 2001.
> > I've uploaded it to http://novalis.org/cases/Moore.html -- let me know
> > if it's garbled in any way.
> > If blueprints are components, then source code definately is.
> So, according to the court, were paper and glue; does this make the use
> of paper and glue illegal in the USA?
Not the use in general, but the provision for the purposes of making a
patented device, apparently.
> Clause (f)(1) states (from the decision):
> "Whoever without authority supplies or causes to be supplied in or from
> the United States all or a substantial portion of the components of a
> patented invention, where such components are uncombined in whole or in
> part, in such manner as to actively induce the combination of such
> components outside of the United States in a manner that would infringe
> the patent if such combination occurred within the United States, shall
> be liable as an infringer."
> While the decision found that blueprints could consititute part of "a
> substantial portion of the components", it was clear that paper and glue
> were also needed. It's not clear to me that blueprints by themselves
> are enough.
Fair enough, but what else do you need to compile a program (and thus
produce a known-infringing binary) other than its source code? A
compiler? Well, Debian's providing that too.
> Furthermore, both the statute and the decision makes reference to
> "actively induc[ing]" the infringing behavior outside the USA. It seems
> clear that we do not actively induce any such behavior; indeed, one must
> expend some effort to get infringing packages built, which would include
> removing roadblocks we actively set up.
(c) doesn't have the "induce" bit -- that's only in (f). But I couldn't
find a (c) case about blueprints (or, indeed, any other 271 case about
blueprints or source code).
> Thus, the case doesn't provide evidence that distributing source code
> for a patented algorithm in itself infringes on a patent. Providing
> tools for compiling the code that are specific to the code, or compiling
> the code and distributing binaries, may be held to be infringing, but we
> don't do that; indeed, we make it more difficult to compile the code in
> an infringing way.
I don't know what, other than source code, could be called "components"
of the patented invention.
> I don't mean to say that removing the LZW code is a bad idea; I think
> it's the safest route. But I am very nervous about the implication that
> uncompiled source code, by itself, can infringe on a patent. It seems
> to imply that I cannot learn about the patent without infringing on it,
> which violates the basic idea of patents: public disclosure in return
> for a temporary monopoly.
There are cases which have ruled that source code or even flow charts
aren't necessary to the disclosure requirements of patents (ask me and
I'll look them up). I know this is not proof, but it is evidence that
source code is more specific than is needed for public disclosure.
Looking at it from a larger viewpoint, the idea that merely distributing
source code and saying, "don't use this" gets around patent law is
fairly silly. The only sane interpretation is that creating source code
is "making" the invention, and that source code is the invention. I
can't see any other interpretation that doesn't lead to absurdity.
GPL Compliance Engineer
Support my work: http://svcs.affero.net/rm.php?r=novalis&p=FSF