[RFC] proposed user doc for nested trees

[Vincent Ladeuil]

Sorry for the delay,
>>>>> "aaron" == Aaron Bentley <aaron at aaronbentley.com> writes:

<snip/>

    >> I think that point is a very important one and should be clearly
    >> defined.

    aaron> Please see
    aaron> http://bazaar-vcs.org/NestedTreesDesign#sub-branches
    aaron> and tell me if there's more I can do to clarify.

Sure, so, the relevant chapter contains may counter-examples,
many negative sentences but the solution you finally retain is
not explicitly defined. So after:

    aaron> When branching from a branch with tree references, bzr should
    aaron> create local branches.

What about adding: "All these local branches will share the
repository of the containing tree" ?

I'm not sure this design is the only possible one, it is, indeed,
enough for a first implementation and we may not need (or can't
implement) another. Yet, it implies that the repository used will
contain the histories of the containing trees and most/all the
histories of the tree references. One may hope that we could
reduce that part in further designs.

       Vincent