I don't understand the hype around CI and that it's supposedly impossible to run something like that without Git, let alone Github. Like sure, a nice interface is fine, but I can do with a simpler one. I don't need a million features, because what is CI (in practice today, not in theory)? It's just a set of commands that run on a remote machine and then the output of those commands is displayed in the browser and it also influences what other commands may or may not run. What exactly is the big deal here? It can probably be built internally if needed and it certainly doesn't need to depend on git so much - git can trigger it via hooks, but that's it?

I think the real problem is we were sold all these complex processes that supposedly deliver better results, while in reality for most people and orgs it's just cargo culting, like with Kubernetes, for example. We can get rid of 90% of them and be just fine. You easily get away without any kind of CI in teams of less than 5-7 people I would argue - just have some sane rules and make everyone follow them (like run unit tests before submitting a PR).

GitHub CI lets me test my package, for free, in all these: https://github.com/ncruces/go-sqlite3/wiki/Support-matrix

Which is actually useful.

> just have some sane rules and make everyone follow them (like run unit tests before submitting a PR)

and thus you discover the value of CI

The big deal is that GitHub provides it for free. Plus it integrated properly into the PR workflow.

Good luck implementing merge queues yourself. As far as I know there are no maintained open source implementations of merge queues. It's definitely not as trivial as you claim.