Interestingly, exactly the feature you mentioned is just being open sourced due to high community demand and it being free in its competitors: https://gitlab.com/gitlab-org/gitlab-ce/issues/34591#note_57...

I based myself on what is available on their "features" page:

https://about.gitlab.com/features/

In any case, this highlights once again the issue: the community will have to put enough pressure on the parent company to release certain improvements as Open Source. And their interests are most likely not aligned.

But doesn't the fact that the feature you mentioned has actually been made open sourced, that those business models might well be aligned? IIRC, GitLab's open sourcing policy is something along the lines of "if it's useful to small or open source projects, we open source it, whereas if it's primarily useful to large enterprises with >100 employees, it goes into enterprise edition". That sounds like a pretty viable way to both have a useful open source project and raise the money required to build and maintain it.

If they don't it can always be forked though. I think that's why they eased up - they realised that GNOME would just fork GitLab if they didn't release this issue.

GNOME already planned to self-host and have enough of a community to maintain a close fork.

I think the bigger picture here is code hosting. You cannot fork that, and TBH the value of moving from Github to gitlab.com is hosting alternative which is not really open-source.