Multiple mirrors sounds like a good idea. I'll be the first one to admit that I don't have enough experience in project management and requirements to pick an alternative and say "I'm 100% sure that's what we should go with." The approach I'm hoping we take with this is staying on GitHub (and possibly mirroring to GitLab and SourceForge) while we test various alternatives and see what works out best for us. I don't think we're in a position now to know ahead of time what the best solution is, and I am hoping we don't have to rush this.New Tobin Paradigm wrote: ↑2020-07-02, 09:36I think whatever is offered should be the best solution available. The absolute junk offered as possibilities in this thread shows a real lack of the practical experence in project management and requirements needed to start transitioning to independent infrastructure. I also think that dumping github completely is a mistake and merely the sole reliance is what should be changed.
Github should be retained as a mirror (and bitbucket and gitlab dot com should also be setup for redundancy) but work should be shifted to a local solution. Something good and supported and most of all exploitable for our needs ie bending whatever it is to support our collective needs.
As I indicated, I am also searching for just a solution for BinOC's needs and if successful in the endevor maybe it will be what is needed collectively and maybe not but what shouldn't happen is a rush job to scramble to change to something unsustainable.
We shall see.
I mean, I've taken a serious look at the most likely alternative everyone is considering, a self-hosted GitLab instance, and if I'm being honest I think it's bad enough that I'm hoping we stay on GitHub if that's our only other choice. GitHub does have a desktop client of some kind (though it is Electron), and besides, one could easily take the perspective that if it comes down to us using Firefox or Chromium to do development because we're missing critical web features, then we have bigger problems than the indignity of using an alternate browser to access GitHub.
This decision absolutely will affect our project for a long time to come, and we shouldn't rush into anything or make a hasty decision given that we have never worked with self-hosted forge software before. This is new territory for us all. One reason why I keep saying "if we switch from GitHub," is because I'm not yet convinced we have a better plan than just using GitHub. It's also why I was saying I think we should choose the best alternative, kept suggesting alternatives, and linked an article discussing forge software as a concept to get people researching the topic and debating alternatives rather than trying to settle on one thing.
The only thing I can offer is that possibly we're looking at this the wrong way. So far we've been trying to find one suite that offers everything via a web interface because that's what we've been accustomed to. But we could also shift to a different workflow, and just move our issue and tracking system off GitHub. We could turn off the ability to make issues/comments on our repo, continue using GitHub for other functions, and move all our existing issues into an independent bug tracking system. We have all the issues separate from the PRs for a good reason, and that fact might come in handy now. Maybe we could make the move in stages, focusing on migrating our issues and bug tracking to a different system first (which is a big job in and of itself), but continue using GitHub as usual for everything else as we replace pieces of what we use it for with more robust alternatives.
Another idea would be to spend some time researching the workflow and software used by other projects of our size and trying to make a list of best practices and features we want our solution to have so we can narrow things down. We kinda started this discussion with no clear parameters, and we all came at it from different directions. Some people think we just need something self-hosted and it doesn't matter what it is, some people think we need a self-hosted clone of GitHub that isn't GitHub, and some of us are worried about long-term maintainability and how far we can scale the infrastructure. We know what we don't want, but the question is, what is it we DO want?