Quoting [EMAIL PROTECTED] ([EMAIL PROTECTED]): > In fact, there are two tests that I know of for determining > derivative-work status: > > 1) If you never saw the original, your work can't be a derivative of it. > > 2) Otherwise, the abstraction-filtration-comparison test applies: we > reduce the elements of the work to their abstract forms, removing all > particulars like the names of variables; we filter out non-copyrightable > elements; and we compare. > > Of course the answer is jurisdiction-dependent.
Yes, I've read about those (though it's been some time, now). > That turns out, as Kevin Renner says, not to be the case. > Micro Star thought their work in issuing a CD full of Duke > Nukem levels didn't infringe FormGen's copyright on the > game, but they were mistaken. See Micro Star v. FormGen, > Inc., 154 F.3d 1107 (9th Cir. 1998). The decision is at > http://cyber.law.harvard.edu/openlaw/DVD/cases/Micro_Star_v_Formgen.html > and is very much worth reading: it contained several eye-openers for me. Yes, I apologise for having momentarily confused the derivative-work issue with that of "viral effects". (I've been having a busy day.) You are of course right that a work, _if_ derivative of another, is encombered by the rights of that work's owner. Whether a source work _is_ a derivative of another is a factual question addressed by courts exactly as you've indicated. But that's not what we were discussing. We were discussing whether a source code work that is otherwise _not_ derivative of another is magically rendered derivative by running the two through a compiler. You seemed to be saying Prof. Moglen claimed this. I was saying it seems highly implausible. Have we miscommunicated? -- license-discuss archive is at http://crynwr.com/cgi-bin/ezmlm-cgi?3