[Monotone-devel] Re: linus talk on git

[Bruce Stephens]

Brian May <address@hidden> writes:

[...]

> As there any references on how policy branches will solve these
> problems?

No.  It hasn't been designed yet.  There are some candidate ideas on
the wiki.

Some bits are easy: for internal purposes, branches are an identifier
(probably a hash), and the branch naming is some separate kind of
mapping between names and branch ids.  (And the same should be true
for keys.)  However, that's just the obvious bit.

>     Bruce> relevant timescale.  (I'm currently not considering
>     Bruce> mercurial because git's MinGW port seems OK, and because I
>     Bruce> don't have a clear understanding of mercurial's local
>     Bruce> branches: they look a bit too new, and not enough like
>     Bruce> git's nicely simple idea.))
>
> How does mercurial cope?

Don't know.  I suspect people mostly use the workspace==branch idea,
and that these named local branches are new enough that they're not
clearly described and probably not that widely used yet.  Makes sense
for many projects, but (mostly because of history) probably doesn't
for us at work.