> I have a repository with 1 project, and some of the directories
> contain code that I would like to maintain separately, so when
> someone
> makes changes, or wants to make a branch, or tag, I don't have to
> take all directories (include those I didn't modify), so, how can I
> do
> that?
>
>
> repo/
> tags/
> branches/
> trunk/
> dirA
> dirB
> (...)
> dirN
>
> What I want is to take dirB out, so when I create a branch
> (because
> of the mades made on dirB, which is another project (and suppose
> dirA,
> dirD, dirF are directories of the same project), only dirB will be
> on
> the new branch, since I didn't modify the other projects. This will
> happen for some other directories (for instance dirE, dirG, ...).
>
> I need to do this because the initial directories tree was not
> well
> planned (as you can easily see), and projects kept being uploaded
> without a proper separation.
>
> I want to copy the entire repository to a second machine, and on
> that machine, properly separate the directories which do not belong
> to
> the original project, this way I have the old machine, and the new
> one, where I can manage permissions easily, since the layout will
> be
> correct.
Just do a dump and load. Are you keeping the old one live, or just as a back up? If you are keeping the old on live you will need to force a new UUID when you do the load, or change it after the load, either way.
I'm not quite sure why you just don't move your folders around in the current repository rather than copy them to a new one before changing your layout?
BOb
Received on 2010-06-29 19:56:09 CEST