On 21/04/2019 18:07, Jeremy Huntwork wrote: > On Sun, Apr 21, 2019 at 11:37 AM Pierre Labastie > <[email protected]> wrote: >> >> I've read a little more about this [1], and wants to summarize here what I >> understand. Note that I've not checked that what I say is valid in countries >> other than US (I've just seen yesterday, when looking at W3m, that a true >> open >> source license is impossible in Japan). jhalfs has been based in US from its >> beginning, so let us consider it is under the US law: >> - All contributors are copyright holders. There's no need to register to be a >> copyright holder, and there is no notion of a minimum contribution to be a >> copyright holder. Actually, all contributors have made substantial >> contributions, so the point about minimal contribution is not relevant here. >> - If there is no license, nobody has right to use, distribute, modify, parts >> he or she has not written, unless given explicit permission! Even other >> contributors have no right to modify what is already written! This is the aim >> of the license to relax such permissions. >> - Jeremy, the initiator of the project has chosen the GPLv2 license, so all >> contributions are under this license. Changing to another license is possible >> only if the new license is compatible with the previous one, unless the >> copyright holders agree to change to an incompatible license. Here, the only >> compatible license is GPLv3. AGPLv3 is not (too restrictive), LGPLv3 is not >> (too permissive), and other common licenses (MIT, Apache, Mozilla) are too >> permissive too. At this point, we have two possibilities: >> - go to GPLv3 (or keep GLPv2, but it is not well suited to modern ways of >> collaborating). >> - Ask the seven contributors whether they accept a more permissive >> license >> (I would push for MIT. Other licenses are not very sensible for >> jhalfs). > > My preference would be to try this first, seeking permission to move > to MIT. If that fails what issue is there with keeping GPLv2? I > believe a move to Github does not really impact the license and I'm > not really a huge fan of GPLv3, although admittedly it's been a while > since I looked at its details. Overall, I think it's just more complex > that it needs to be. I like the simplicity of MIT or BSD licenses. > >> - Gihub has two types of repo: >> - private, means a few collaborators (maximum of 4 with free github) can >> access the repository, but it is not visible to anybody else >> - public, means it is visible to anybody, and anybody can be given >> commits >> right, but there are again to possibilities: >> - owned by an individual, who has all the administrative rights. >> - owned by an organization. Means there may be several owners, which >> may give various rights to users (administration, commit, etc, I've >> not read it in full yet) > > Private would make it hard to collaborate and I think kind of defeats > the purpose. Given the history of ALFS, I'd say an organization (you > can create one and invite others to be admins) makes the most sense.
OK, I'll send a mail to all the other five contributors (I think I can consider having Jeremy's agreement, and mine). I agree with creating an organization. Ideas for name? Pierre -- http://lists.linuxfromscratch.org/listinfo/alfs-discuss FAQ: http://www.linuxfromscratch.org/faq/ Unsubscribe: See the above information page
