Hi Pierre, Pierre Neidhardt writes: > I haven't really followed the issue, I have, very carefully ;-) > so I couldn't say whether the decision taken by the core maintainers > was right or not. From my point of view it was /but/ this is *not* relevant: what's relevant here is that /if/ we trust Guix maintainers (I do) when they give commit access rights to people, whe /have/ to trust them when they revoke those rights. We /should/ disccus /if/ the rules and best practices to have and maintain the commit acces are well documented: please make proposals (patches wellcome :-) ) but please we have to keep trusting Guix maintainers (that is a collective of very competent people). [...] >> I am not a core maintainer, but it should be obvious that core >> maintainers would not take a decision to revoke commit rights lightly. > > I trust that it is the case, but being the devil's advocate, Please don't: «Why the Devil's Advocate Doesn't Help Reach the Truth» https://www.gnu.org/philosophy/devils-advocate.html --8<---------------cut here---------------start------------->8--- The devil achieves that by twisting my words: presenting a misleading context in which my words appear to mean something other than what I intended. --8<---------------cut here---------------end--------------->8--- ;-) [...] > Another question one could ask: why just the core maintainers > actually? Shouldn't everyone be involved? Maybe the right answer is > "no" here, and if so, I believe we should explain why in the community > guidelines. Guix is a GNU project and AFAIU GNU project management is well documented: https://www.gnu.org/gnu/gnu-structure.html I don't know if Guix project needs specific «GNU Guix structure» documentation but /if/ the answer is yes it should complement the official GNU one, not replace it, IMHO. BTW I see Guix contributors with commit access as "package maintanance assistants" delegated by maintainers to make some technical decisions: --8<---------------cut here---------------start------------->8--- The maintainers of a package often recruit others to contribute to its development, and delegate some technical decisions to them. However, the maintainers retain authority over the whole of the package so they can carry out their responsibility to the GNU Project. --8<---------------cut here---------------end--------------->8--- Please we should always consider that GNU maintainers are the persons that carry out the responsibility to the GNU project, not contributors with commit access. Maybe the contributing section of Guix manual should mention it and link the relevant GNU project's documents: do you think it'd be useful? > Lest the community present an image where a few would benefit from > arbitrary privileges. ...or seen from /the other side of the moon/: a few carry out the precious work to be /responsible/ to do a good, practical job of developing and maintaining Guix according to the GNU project's mission and general decision. If you want call it /arbitrary privilege/ but I have a different point of view :-D The "community" (whatever this means) should acknowledge that contributing also means to be responsible toward other users of free software: this needs competence in the specific matter (also domain specific), discipline (i.e. properly documenting changes in commit messages) and commitment to a set of common shared rules (documented in Guix and GNU project manual). [...] > Last, maybe a more important question: if core maintainers are > entrusted to take executive decisions about the community members, > what about executive decisions about the core maintainers themselves? Maintainers are appointed by the GNU project. > Are there such provisions? Example: what if a core maintainers > misbehaves? Can they see there privileges revoked? How? Is this > documented? «GNU Software Evaluation» https://www.gnu.org/help/evaluation.html#whatmeans Does this answer your question? [...] Happy hacking! Gio' -- Giovanni Biscuolo Xelera IT Infrastructures