it's an ongoing frustration that "public/private" visibility can only be controlled at the repo level.
would probably not be too hard to write a few scripts/workflows to facilitate making a private clone behave as though it were hosting private branches of a public repo.
- headless clone "mirror"
- scheduled github synching workflows
- "publish" script publishes a branch to a public fork of the target repo and creates a PR to the origin repo
- PUBLIC - origin
- PUBLIC - fork (upstream=origin)
- PRIVATE - mirror (upstream=fork or upstream=origin)
Q: should POC start from a private repo we want to publish, or forking a "private branch" off of an already public repo?
- ultimately want to support both of these, but need to pick one to start with.
- probably simplest to assume the public code already exists and we're creating a fork.
- the specific use case I want to address has both already existing in a disconnected fashion
- ... need to start somehwere. let's start with adding a "private branch" to an already public repository
to do: come up with a better name so I can stop using quotes everywhere.
https://stackoverflow.com/questions/3959924/whats-the-difference-between-git-clone-mirror-and-git-clone-bare https://docs.gitlab.com/ee/user/project/repository/mirror/