On 4/14/05, Humberto Massa <[EMAIL PROTECTED]> wrote: > David Schwartz wrote: [snip] > > There are court cases on point that definitely > >disagree with you, for example Mirage Editions, Inv. v. > >Albuquerque ART (cutting a picture out of a book creates a > >derivative work). Also National Football League v. TVRadio > >Now (embedding someone else's broadcast with your > >advertisements through an automated process creates a > >derivative work). > > The embedding was not made by a fully automated process, was > it? Didn't someone had to create the advertisements, with the > purpose to be presented embedded in the broadcast? I suspect > -- without looking at the case files at the moment -- that > there was the creation of the derivative works...
I don't think the Mirage Editions court cared whether the process was fully automated or not. They rejected a defense based on the "first sale" doctrine (once I buy a copy of a book, copyright law has nothing to say about the conditions under which I may resell the physical object) because Albuquerque A. R. T. "recast or transformed" fragments (individual pictures) from the book prior to resale. Implicit in the court's logic is the fact that the copyright holder faced, not the ordinary effect of the used book market on new book sales, but the competition of Albuquerque A. R. T.'s tiles in the market for art reproductions. The software equivalent would be subscribing to MSDN (a subscription to many Microsoft product releases for development and test purposes) and then claiming "first sale" justification for selling the individual CDs as retail software. Separate from licensing agreement considerations, I think a court would find that the "first sale" defense is inappropriate in this situation. In my opinion (IANAL), the Mirage Editions court ought to have given less weight to the "derivative work" angle and more to "the doctrine of first sale isn't intended as a cover for deliberate misappropriation for competitive purposes". In any case, the case can't be stretched to cover intrinsically uncopyrightable fragments, which is how recent US courts have approached software interfaces required for (hardware, software, and even wetware) interoperability. The grant of preliminary injunction in the TVRadioNow case ( http://www.mpaa.org/Press/iCrave_Findings.htm ) doesn't appear to have involved "derivative works" at all, nor to have hinged on (or even mentioned) ad substitution. The plaintiffs proposed a set of findings of fact and conclusions of law, focused on the right of "making available" (Berne convention) / "public performance" (17 USC), which were adopted by the court as grounds for the injunction. Commentators (including the MPAA and the WIPO) have said all sorts of stupid things about this case, but as a legal precedent with respect to the GPL it seems irrelevant. > > I think it would make a lot of sense if courts held > >that compiling and linking are analogous to format changes > >(like converting an audio-visual work from DVD to VHS). This > > Our (.br) courts do. I don't know (I'd have to read the cases > you cited) why did those courts ignored the intellectual > novelty requirement of a derivative work, but I'll look into > it. Citations supporting this assertion would be very interesting. But even so, I doubt that linking would be held to create a new _derivative_ work rather than a _collection_ in a case that actually hinged on this point, especially if the collection is of such a nature as to contain no fresh creative intent. Personally, I think the notion that there is a legal (as opposed to an engineering or moral) objection to supplying firmware blobs in conjunction with GPL drivers (compiled in or not) is almost as silly as claiming to be offering the hex blob as an intrinsic portion of the driver's source code under the GPL. The vendor would almost certainly (IANAL) be enjoined from objecting to its being copied, with or without modification, by anyone for any purpose once they've encouraged an employee, contractor, or regular correspondent to plop it into Linus's tree. But I doubt that its inclusion in a GPL driver could be used to pry its source code out of the vendor. Cheers, - Michael