On Wed, 21 Dec 2022 14:36:31 +0000 drkhsh <[email protected]> wrote: Dear Aaron,
> So, before this discussion becomes untechnical, here are some facts > from a quick research. > > https://www.copyright.gov/title17/92chap4.html#408 > > "(2) in the case of a work other than an anonymous or pseudonymous > work, the name and nationality or domicile of the author or authors, > and, if one or more of the authors is dead, the dates of their deaths; > > (3) if the work is anonymous or pseudonymous, the nationality or > domicile of the author or authors;" > > For me that means that publishing pseudonymous copyrighted work under > any license should be fine as long as the license does not explicitly > mention it? you are citing US copyright law and it's a whole can of worms across the world. I could go on citing EU, Russian, etc. (all signers of the Berne convention) laws, which are lenient or dismissive in regard to pseudonymous attributions, but always at least require unique identification of a pseudonym, which can always be a matter of dispute. If a license is invalid in a certain country's jurisdiction (e.g. if pseudonyms are used even though not allowed), the code ends up being "all rights reserved" and thus not satisfying the OSI license criteria. Even if pseudonyms were allowed by the Berne convention (didn't check), it would probably at least require unique identifiability of the pseudonym, casting doubt at the overall license text. I know many people and companies who are very careful about only building their software using components with watertight licenses. Regarding the pseudoynm uniqueness, I think that "pseudonym" in the law's sense is very thinly stretched in regard to arbitrary web-nicknames, and the only reason, I think, it's included in some laws is that when an author writes a book under pseudonym the book doesn't immediately go into the public domain. We could discuss this forever, but this ends up being territory where one would have to ask a judge if a pseudonym is unique enough or not. Having real names only in a license has other, more practical, advantages, though: You actually have the chance to reach out to people even years after the software release, and I've had one positive experience with this a few years ago. There is simply no chance if you just have a nickname with a throwaway-e-mail-address, e.g. "PBC <[email protected]>", to ever reach out to this person in most cases after just a few years. Reaching out could be regarding a relicensing (e.g. ISC/MIT -> GPL or the other way round), technical inquiry or simply to invite them to something. I'm amazed about how many people are scared of putting their names on their works; in a sense it reduces the software's trustworthiness, may it only be by a subjective factor. With best regards Laslo
