Hi, debian-science. debian-legal had a query about a program which had a citation requirement in its licence. See below. What's our usual approach ?
Boud Roukema writes ("does MUSIC (cosmology package) qualify as free under DFSG?"): > I would like to use the MUSIC cosmological initial conditions software > https://bitbucket.org/ohahn/music ... > ii. does not allow modification You are right. This is probably a mistake by the upstream authors. You should see if you can get them to fix it. If they won't then unfortunately, unless we can find some pretty good other evidence of their actual intent, this can't be in Debian. > iii. does not allow distribution [conditions on distribution are > listed, but that could be interpreted to mean that *if* you obtain > permission from the author to distribute, *then* this is a > constraint/reminder about conditions on the private permission that > you have obtained] It could. I think it's an odd interpretation, and a court would probably decide that the copyright holders intended to permit distribution (particularly, since they have put the thing on a public git hosting service). But again, you should ask them to fix it, since you're asking them to clarify anyway. > iv. does not allow distribution of modified copies See (ii). > v. requires obligatory citation of the software and research paper > > On point v: the GPL forbids this: > https://www.gnu.org/licenses/gpl-faq.html#RequireCitation > "No, this is not permitted under the terms of the GPL." That a licence is not GPL-compatible is unfortunate, but does not make it non-free. > On the other hand, the same GPL FAQ seems to imply that a citation > requirement is legally invalid. Does it matter if a licence has a > non-enforceable (illegal) requirement that people will generally > follow voluntarily (and under academic ethics rather than legal > obligation)? Or rather: would this be accepted under DFSG? I don't know what Debian's position is on this. I think the GPL FAQ entry does not mean that this licence is unenforceable. If it is enforceable, it may or may not be a problem. Personally I don't have a problem with it, but it's not my opinion that counts - I'm not one of the ftpmasters. I know that in some other cases upstreams have been persuaded to change such licence conditions into non-binding imprecations. In this case that would make the resulting software GPL-compatible, which would be very nice. Perhaps upstream can be persuaded. And perhaps debian-science can come up with some examples to help persuade. BTW, I am very sympathetic to the underlying motives for these kind of licence terms. There is a big problem with lack of funding for scientific software. See for example https://arxiv.org/abs/1610.03159 _The Astropy Problem_ for a description of one instance of the general class of problem. Regards, Ian.