[ANNOUNCE] bzr-gardener: A new way to manage lots of branches and working trees

[Russ Brown]

On Thu, Sep 16, 2010 at 9:57 AM, Vincent Ladeuil <v.ladeuil+lp at free.fr> wrote:
>>>>>> Russ Brown <pickscrape at gmail.com> writes:
>
>    > This looks really useful! I have a couple of questions though.
>    > How does it determine "merged"?
>
> By looking if the current tip is part of the ancestry of the merge
> url. The merge url is either the submit branch or the parent branch.
>
>    > For me, "merged" would mean the branch has been merged to a
>    > specific other branch (i.e. trunk, or similarly purposed).
>
> Yes, that's the idea and generally that's the parent your forked from or
> the branch you submit your patches to. Both are remembered by bzr so I
> use them.
>
>    > It doesn't seem to be doing that though: it's marking plenty of
>    > branches as "merged" that I don't consider to be merged at all.
>
> Can you post the output of 'bzr info -v' for such a branch ? Does it
> mention either a submit branch or a parent branch ?
>

I think in my case it is probably a case of it choosing the "wrong"
option out of the submit and parent branch. In my case it was checking
against "submit", when "parent" would have been more sensible. I'll
see what falls out of bug 641043 before filing anything about it
though.

> The plugin currently missed the ability to define a "kind" of branch for
> which more precise checks can be made so all feedback about such strange
> results are warmly welcome, you can even file a bug for it.
>
>    > Also, for checkouts where the branch it is bound to has
>    > disappeared, it is erroring (exception: Not a branch). Maybe it
>    > would be cleaner to mark these as something like "orphaned"?
>
> I haven't worked on checkouts yet, but this seems like a very good
> idea. Care to file a bug for it too ?
>

bug 641329 raised for this.

>      Vincent
>



-- 

Russ