Hyman Rosen <[EMAIL PROTECTED]> writes: >Again, I assert that a program written to dynamically link with >a GPLed library, which requires that library for its operation, >may be distributed on any terms its author chooses. The FSF says >that such a program must be distributed under the GPL. Wondering >about how the GPLed library gets to the user's machine isn't >germane as long as it gets there as authorized by the GPL.
OK, using your example, suppose the author of HyProg distributes it along with libGNU on a CD-ROM. Mere aggregation of both together on the CD-ROM doesn't make HyProg subject to the GPL. So far so good. Now the end-user gets the CD-ROM, follows the instructions, and HyProg is now dynamically loaded and linked with libGNU. Who caused libGNU to load into memory? The end-user or the author? I submit that both did, because each is a link in the chain without which the final load of libGNU would not have occurred. But whom will the courts hold responsible in a legal sense? I submit that this question has not yet been answered in the case law. If the author is held responsible, then the author must comply with the GPL throughout the sequence. If the author is distributing libGNU and HyProg together such that they will be used together, and HyProg requires libGNU, then the author has ventured beyond mere aggregation. The HyProg + libGNU combination is now a single work, because both pieces are loaded together when the author's instructions are followed. Is this a derivative work or a mere collection? Is an automobile a mere collection of engine, wheels, and other parts? -- Rahul http://rahul.rahul.net/ _______________________________________________ gnu-misc-discuss mailing list gnu-misc-discuss@gnu.org http://lists.gnu.org/mailman/listinfo/gnu-misc-discuss