Please add LICENSE file

[soupglasses]

Hi i just noticed this awesome project does not seem to have any LICENSE file to declare its status of use.

Please do consider adding one to clarify its possible use. NixOS and Nixpkgs typically follow MIT licensing if you want to follow a norm. :)

[Artturin]

@guibou Please take a look.

A similar case in flake-compat edolstra/flake-compat#23

[guibou]

There is a merge request here: #174

I've asked a question recently about who should receive the "main" copyright on the header of the licence (I'm not sure I need to get my name on top of the licence, considering that I'm not doing anything for nixgl since years)

[soupglasses]

I think the much larger issue is how to legally do it, as there exists 26 authors you'd need to get permission from, or otherwise remove from the commit tree. Sadly i don't think this is as simple as merging in a LICENSE file, at least in a legal sense.

[guibou]

I think the much larger issue is how to legally do it, as there exists 26 authors you'd need to get permission from, or otherwise remove from the commit tree. Sadly i don't think this is as simple as merging in a LICENSE file, at least in a legal sense.

That's the problem. nixGL was clearly initially published as a "hack" in a "do whatever you want with that, I don't care" and unfortunately, it was used and contributed on by users.

I don't have much motivation to work on that, and what I don't really want is to risk any complain or legal actions.

I failed here as an opensource developer.

Do you think if we list all the commiters and people who interact in github in the copyright it would be ok? I mean, it is easier to "remove" someone later from the contributor list if they request it than to add someone we forgot?

[aware70]

@guibou I think you would still need their permission, because you are licensing their work.

I think the safest way to proceed is to tag them all as reviewers on the license change MR and get their explicit consent.

It may take a while for them to do it, but that seems best (I am not a lawyer).

[soupglasses]

There is also the "creative works" requirement. So from my I am not a lawyer perspective, shouldn't apply for very small commits like typos that anyone could quickly spot and do without any other real way to create such a PR. Mozilla also says "95% acceptance rate" is enough for a relicense as well if nobody objects, but this i dont think has been ever tried in court? https://blogs.fsfe.org/ciaran/?p=58