Comment 2 for bug 237439

Hmm, maybe it's my lack of understanding then. In my head I see a shared repository as being a place for branches to be stored underneath, not a branch itself, and that is certainly how we intend to use it.

I can however also see one of my colleagues accidentally pushing a branch over the shared repository itself, instead of underneath it, and I'm uncertain of what effect that would have on the shared repository itself. If the answer to that is zero, then I suppose it doesn't matter.

The nested branches example in the page given intrigues me. Is the history for all of those branches stored in the shared repository at the root? Does that happen in all cases (i.e. bzr always scans for the first shared repository it finds in its parent directories), or does it work by creating branches within branches which all refer to the same root shared repository?