Sam Hartman <[EMAIL PROTECTED]>: > C> Many packages contain preprints or reprints of academic papers > C> as part of their documentation. In many cases, there is no > C> ``source'' available for these documents -- they are > C> distributed as PostScript or PDF files. > > One case that seems fairly clear to me is cases where the upstream > doesn't have the source either. If the upstream would be stuck > editing the ps or pdf if they wanted to modify the document, then that > document is its own source code at least under the GPL definition.
I agree that the PS or PDF should count as source if the original source has been lost, but upstream not having the original source isn't quite strong enough. I would guess that in many cases the upstream author of the software doens't have the document source, but someone else does. I would also guess that in most cases the availability of source is irrelevant, because the academic paper isn't available under a DFSG-free licence anyway; most authors of academic papers don't want other people distributing modified versions of them. This isn't a serious restriction, because no one would want to do that anyway, but it means that academic papers are generally not DFSG-free. Edmund
