Centralized storage in bzr

[Martin Pool]

On 15 Jun 2005, William Dod? <wilk-ml at flibuste.net> wrote:
> >
> > Hmm, I wonder if it would make sense to have a 'bzr release' command
> > that would store the branch in the central storage and remove the
> > working directory.
> 
> I wanted to ask about a 'bzr backup my-backup-0.x.tgz' with all the .bzr
> data and the versioned files... Specialy since there is now somes
> cached data inside (we should have an easy way to remove/identifiy them)
> But maybe it's better to don't mix too much revision control and backup,
> it's not the same job.

What I'd suggest you do is bzr branch to a temporary directory, then
tar that up.  This is basically what John's daily-snapshot scripts do.

This does mean you'll get a complete working copy as well as the
history.  For the current simple storage the history will tend to be
much larger so it doesn't matter so much.

In the future it might be nice to have a branch with no working copy
instantiated.  On the other hand for backups it's good to have a bit
of extra redundancy, and having the working copy means at least you
can get that out even if (because of alien invasion) you can't run bzr
anymore.

-- 
Martin