They tried that by making a bit available with a remote cloud builder for Bazel. It failed for some reason and they pulled it.

I think building something that scales for one big repo is just a completely different problem than making it scale for a lot of small repos.

Long term projects like this don't get any attention because the chance of getting a promotion from it are almost nil.

And after the layoffs, it's pretty clear that no matter how hard you work, you can get fired so what's the point in dedicating your career to something like this?

Bazel is not failing in the open source world though

I think that maintaining a hosted service has significantly higher fixed costs than maintaining an open source project whose users are responsible for deploying it themselves. So a higher degree of adoption would be necessary to justify it.

Dont get me wrong, I love Bazel. But tooling for it to build multi-language multi-billion line monorepos doesn’t exist outside big companies.