Re: RES: What makes software copyrightable anyway?
On 5/18/05, Raul Miller <email@example.com> wrote:
> As usual, I don't know what your point is.
> If I read you right, you're claiming that I mis-read my own question --
> that I really wasn't asking for clarifcation on why you asserted that
> the GPL is not a license.
OK, let's clarify that. There is a particular legal use of the word
"license", as in the phrase "scope of the license", which refers
specifically to an individual provision in a contract that says what
rights the copyright holder (or, in other settings, patent holder,
etc.) is offering to the licensee. Under the 1976 Copyright Act,
there are a variety of rights (to copy, to adapt, to translate, to
write sequels, to anthologize, etc.) that are divisible when issuing
If the license granted is limited to the making of copies, and the
licensee exceeds its scope by modifying the work (and thereby creating
a derivative work), then this "conduct outside the scope of the
license" is not part of the contractual relationship between the
parties, and the licensor has a cause of action under copyright law
(the tort of copyright infringement). This is the _only_ "shortcut"
of which I am aware under US law that allows a plaintiff to prevail
under a copyright law theory while there is a valid, unterminated
contract licensing rights to the same work to the defendant.
This "scope of license" construction does not involve any fine
judgments about whether the licensee's return performance is up to
snuff. Claiming that the entire GPL is a "license" in this technical
sense is erroneous. You don't have to believe me on this, you can
read SOS v. Payday yourself, and see how it was applied in Sun v.
The everyday, interchangeable use of "contract", "agreement",
"license", "license agreement", "terms and conditions", etc. spills
over into legal opinions as well, of course; but I split this
particular hair at this particular juncture because I don't want to go
down the rathole of misinterpreting the phrase "scope of license" to
refer to the entirety of the GPL. OK?
> If I read you right, you think that my citation of this case involving
> colorization was presented as something more than an example of a case
> based on something other than contract law.
You keep asserting that "the GPL is not a contract", apparently
meaning that principles of contract law in a given jurisdiction
shouldn't be used to construe its legal meaning. You haven't advanced
any other theory under which it has any legal relevance; but you keep
(IMHO) grasping at straws that someone has told you are connected to a
court decision in which the terms of an agreement were analyzed using
some other theory. Heritiers Huston is the latest straw in the
series, about which you said:
Huston v. La Cinq Cass. civ. 1re (28 May 1991). is an
example of a court decision that applied some legal theory
other than contract in order to analyze the scope and
effect of a license.
That's a misunderstanding of the import of Heritiers Huston. The
terms of the license between John Huston and the original producer of
"The Asphalt Jungle" were brought into evidence and strongly urged by
Turner Corporation, but were ultimately irrelevant to the judgment in
the case. The plaintiffs prevailed on a "droits morals de l'auteur"
theory, founded in the French equivalent of copyright law -- a
statutory override on what rights it is possible for an author to
contract away, just like the termination provisions in 17 USC 203.
> As for the scope of license issue... you seem to be
> referring to a principle used in patent law, and a principle
> which people think should be incorporated into contract law.
Sigh. If you're going to read one court decision, how about making it
Sun v. Microsoft? The application of "scope of license" to copyright
licenses is staring you in the face in the very paragraphs that we
have most recently been wrangling over.
> As near as I can tell, you're asserting that in the context of a single
> instance of a single work the GPL can be factored into multiple licenses
> between a single copyright holder and a single licensee. And, you seem
> to be asserting that because this is the case that it's not correct to
> refer to the GPL as a license between these two parties, but instead
> it is oly correct to refer to the GPL as an offer of contract between
> these two parties.
> Have I got that right?
That's not what I'm asserting, but note that under some circumstances
it can be true. See the discussion of "contractual covenants" in the
Sun v. Microsoft saga.
> If not, PLEASE simply state your point.
> If so, I think what you're really saying is that when the GPL uses the
> phrase "this license", it does not refer to the "GNU General Public
> License" but to some particular instance of its application.
> If that's not what you're thinking, could you please state your ideas
> more simply?
> In other words, I think you're saying that a court would conclude that
> section 4 of the GPL is meaningless -- that the GPL granted the
> licensee multiple licenses and that only some of them terminated.
> Again, if this is not what you're saying, please make your point more
I'm not saying any of the things you suggest in this last chunk. I'm
saying that the "scope of the license[s]" granted in the GPL is to be
found by construing only those terms defining the rights being made
available to the licensee, not those terms binding the licensee to
return conduct. That scope clearly includes copying, modifying,
translating, anthologizing, and in fact pretty much every right that
is reserved to a copyright holder and can legally be offered under
contract. So there is little or no prospect of any "GPL infringement"
lawsuit in which the plaintiff doesn't have to prove material breach
of contract under the most unfavorable construction the defendant can