To be honest, I didn't really have energy to read all of the thread, by skimming it,

I would like to add my thoughts w.r.t. the path forward.

I really like the idea of having a RedHat/Fedora/CentOS provided an supported GitLab.

But as we plan to ahe gitlab as the alternative to pagure as both gitforge and distgit,

this sounds like two projects, with two distinct pilots and testing periods and re-evaluation.

What I really want to be mindful are sunk-cost fallacies and unknown-unknowns.

So we need to be able to say "we tried but we need more resources" or even

"we tried but it is not working out and we are pulling the plug".

We might get to a point where we'd see that GitLab is a great forge, but we drastically underestimated resources

to actually support migration to use it as a new dist-git.

I have been through a plan to replace an established but under-staffed/under-developed tool at RedHat.

As a QE at RedHat, I have been through a switch from our TestCaseManagementSystem to Polarion.