Re: What's cooking in git.git (Jul 2013, #07; Sun, 21)

[Jens Lehmann <Jens.Lehmann@web.de>]

Am 22.07.2013 22:47, schrieb Jens Lehmann:
> Am 22.07.2013 09:48, schrieb Duy Nguyen:
>> On Mon, Jul 22, 2013 at 2:32 PM, Jens Lehmann <Jens.Lehmann@web.de> wrote:
>>> Am 22.07.2013 08:57, schrieb Junio C Hamano:
>>>> * jl/submodule-mv (2013-04-23) 5 commits
>>>>  . submodule.c: duplicate real_path's return value
>>>>  . rm: delete .gitmodules entry of submodules removed from the work tree
>>>>  . Teach mv to update the path entry in .gitmodules for moved submodules
>>>>  . Teach mv to move submodules using a gitfile
>>>>  . Teach mv to move submodules together with their work trees
>>>>
>>>>  "git mv A B" when moving a submodule A does "the right thing",
>>>>  inclusing relocating its working tree and adjusting the paths in
>>>>  the .gitmodules file.
>>>>
>>>>  Ejected from 'pu', as it conflicts with nd/magic-pathspec.
>>>
>>> So I'll base my upcoming re-roll on pu, right?
>>
>> The conflicted part is the use of common_prefix. I think you might be
>> able to avoid the conflict by using quote.c:path_relative() instead of
>> common_prefix() and prepending "../" manually. Or not, I did not read
>> path_relative() carefully, nor your connect_work_tree_and_git_dir().
> 
> Thanks for the pointers, I'll look into that.

Yup, relative_path() seems to be the solution here (and makes the
connect_work_tree_and_git_dir() function much shorter :-).

What worries me is that even though t7001 breaks because of this
conflict (just like it should) when run inside the t directory by
itself, the prove and normal test runs did not report any failures.
I have no idea what is going on here ...