Re: Where to put Open Transport Tycoon (openttd)
On 5/16/05, Sean Kellogg <skellogg@u.washington.edu> wrote:
> On Monday 16 May 2005 05:31 pm, Michael K. Edwards wrote:
> > Raul, a work made by collecting X, Y, and Z is not a derivative work
> > of X. Not even if the "selection and arrangement" involved is
> > original enough to be copyrightable, and a fortiori if it is not (as
> > in the case of Quagga + Net-SNMP + libssl, an obvious combination
> > (given the text of each) if ever I saw one). Assert it to your dying
> > day, that's still not what the phrase "derivative work" means in
> > modern jurisprudence, anywhere in the world.
>
> I apologize for appearing stupid, having failed to keep abreast on current
> debian-legal discussions, but I can't help but think there is something wrong
> with the paragraph above. Why exactly is a collection of X, Y, and Z not a
> derivative work of the three? Does this mean I can buy a three books (X, Y,
> and Z) tear off the bindings, rebind them into one book XYZ and resell at a
> profit? The questions surrounding what is and is not a derivative work are
> quite complex and are not made any easier by the confusing definition in the
> Copyright Act.
The 17 USC 101 (1976 and later) definition isn't that bad, and the one
in the Berne Convention is even clearer. Derivative works are
sequels, translations, adaptations to the silver screen, that sort of
thing. Collective works are anthologies and encyclopedias and the
Color-Coordinated Leather-Bound 2005 Main and Alternate Selections of
the Book-of-the-Month Club.
These two categories exist for different historical reasons and have
separate legislative and judicial histories (note that there is no
definition of "derivative works" in the 1909 Copyright Act), so the
presence of separate definitions in today's 17 USC 101 is no surprise.
The statute doesn't need to say they're technically disjoint
categories (which is what all that silly symbolic math is about);
almost no judge would confuse them if a case's outcome actually
depended on the distinction, and if one did, that's what appeals
courts are for.
Authorization to publish copies is just that -- authorization to copy
-- and it takes explicit authorization (a separate license "scope") to
do any of the other things. (In some cases this license may be
implied through conduct; it's still a creature of contract law,
though, as outlined in Effects Associates v. Cohen, and this is long
since baked into Nimmer.) Sane people authorize something specific
instead of playing games with the legal categories; the GPL doesn't do
the sane thing because it pretends not to be governed by contract law,
and as a result the drafter is hoist upon his own petard.
> The question, it seems, is whether the compelation is "a work based upon one
> or more preexisting works." Certainly my compelation book XYZ is based on
> the preexisting works of X, Y, and Z... but that doesn't help us figure out
> if its a "work" in the eyes of the copyright statute. Generally this
> requires some showing of original authorship... but your claim is even if
> the "selection and arrangement" is original enough to be copyrightable, its
> still not a derivative work. But that doesn't seem right... Book XYZ is
> both "a work" of original authorship AND "based upon one or more preexisting
> works."
That is not the question. The question is whether it was created from
the original work[s] by "recasting, transforming, or adapting" them
(derivative work) on one hand, or by "assembling into a collective
whole" (collective work) -- or, if you like, "selecting, coordinating,
or arranging" them (compilation; and if the originals are themselves
copyrightable, collective work) on the other hand. The Berne
Convention doesn't go to the trouble of putting "one or more original
works" into the definition of "derivative works", which is just
additional evidence that historically collections are not a flavor of
derivative work.
> I'll readily admit to being inexperienced in this area. I'm just finishing up
> my second year in law school... so I've got lots more to learn and am happy
> to be proved wrong.
I think I might have visited a law school once. :-) IANAL.
> p.s. PLEASE don't cite cases to make your point... reading a whole case takes
> a lot more time that just explaining the logic used by the Court.
Fair enough. I think you will find extremely few cases that hinge
upon the distinction anyway. The only purpose for which it really
matters that they are separate legal categories is when somebody
writes a generic offer of contract, licensing the creation and
distribution of "derivative works under copyright law" in return for
some specific performance, instead of a specific license to translate
"American Sphinx" into Klingon in return for $x per copy sold. But
when someone does so -- as the drafter of the GPL has done -- I think
a court should have no difficulty in acknowledging the licensee's
right to have it construed narrowly if he or she so requests.
Cheers,
- Michael
Reply to: