The simple solution for that was to just move to gitlab instead of a self-hosted instance … but there are some tradeoffs (though if no work is going to go into ‘fixing’ the manjaro gitlab instance … the tradeoffs are a major net win for the move)
Or ‘accepted’ users could be allowed to fork code and make proper PRs on the self-hosted instance.
Or …
There are probably a thousand other ways it could be done … but none of them have been tried … let alone seem to be seriously considered.
Another case in point - I still have a working manjaro-stable downgrade server [and patched downgrade utility] … thats built and working … that I pay for … its been going for over a month now … and I still havent gotten any serious feedback.
OK. But I never even asked for ‘open to public’ … we have dozens of ‘vetted’ accounts on the instance already … but they have no ability to do anything.
Mine for example has access to artwork … I can edit or fork or whatever those … but no other section.
So … if it is going to be managed this way … it actually needs to be managed. Thats the issue.
? if you can fork , you have all ! in all project main page exists a link “ask access” - after we can create new branch in this project (no fork to create) and send PR on “master”
No my point is I cant fork or do any of those things.
I now have access to only artwork group.
(and I dont know when that happened … I was never notified … and since it has always worked like this I never check the account until recently)
And again … this is after I have been around for years and contributed some of that code and bothered people … and this is the situation I am in after all that.
Someone coming along today will have even less luck trying to get something done.
If things are like I read in this post, seems to me that the gitlab own instance is like a poisoned gift. It looks great from outside but it’s affecting the project badly. The own instance is somehow killing the community feedback
That’s not really an argument though.
There is tons of donation money available to pay those.
I guess the main problem is that the core team has too much work in general atm and thus focus on the core things. Reviewing, testing and accepting the code contributions costs time… When it is on gitlab and everyone can fork and contribute, this will be even more of course.
So basically the core team has to grow imho. Not sure who can do what atm, but maybe some team members should get some more permissions in order to relieve the core team from some tasks.
(Which is difficult to decide for the core team of course. Who can you trust? Who has the same opinion and vision as the current team? -> Who can you give those permissions…)
The license we have with gitlab grants many extra features, but we are using hardly any of them. So moving to gitlab.com would not have that many adverse effects.
I have been evaluating and testing some options, but it seems that
There is no feasible way to enable easy forking with the self hosted gitlab.
Mirroring the repos to gitlab or github would cause too much fragmentation.
So, I’m also in favor of moving to publicly hosted git solution. Github is more popular, but since gitlab and github allow sso between each other, it is not a big deal. Gitlab offers repo groups, so it would be easier to migrate there and keep the current project structure. And some people don’t like that m$ owns github, so that’s another point in favor of gitlab.
it should not be exaggerated, at github time there were very few PR external to the team. The problem here is that we have the impression (false) that manjaro doesn’t want contributors.
it looks like an obstacle course and undocumented to make things easier
Manjaro (the team) definately wants and supports contributions, no doubt about that. At least that is my experience.
Looking at things like @cfinnberg and @cscs want to contribute and they never got feedback looks to me as if the team did not yet have the time yet to review those things (or at least say: no we don’t implement that because…). Maybe also just going overlooked in the tons of requests they get. -> hence might indeed think they would not give a shyte.
Problem is, it is often unclear who is responsible for what. I have gotten a lot of notifications about people requesting access here and there, but I only grant access to manjaro-architect and the PKGBUILDs that I maintain. If people request access to something I don’t work with, I do nothing about it. I don’t know what others do.
I was checking if I can grant access to manjaro-tools repo to @cfinnberg, but turns out I don’t have access there either.
This please.
I would love for Manjaro to stay with GitLab even if it about moving to GitLab dot com
XFCE and Arch Linux among others in the Linux world uses GitLab for a reason. Albeit them being self-hosted instances. Ideally GitHub should be the last option.
Some of those things could be opened up without opening the registration process.
Allow some more movement within gitlab … but still gatekeep new accounts.
If people request access to something I don’t work with, I do nothing about it. I don’t know what others do.
and snippets are clear …