On Tue, Nov 9, 2010 at 3:25 AM, Junio C Hamano <gitster@xxxxxxxxx> wrote: > Nguyen Thai Ngoc Duy <pclouds@xxxxxxxxx> writes: > >> On Mon, Nov 08, 2010 at 09:32:03AM +0100, Michael J Gruber wrote: >>> Add a test for alias expansion in a subdirectory of the worktree. >>> >>> Signed-off-by: Michael J Gruber <git@xxxxxxxxxxxxxxxxxxxx> >>> >>> --- >>> Â Â 65f3a9e (Remove all logic from get_git_work_tree(), 2010-11-01) breaks this test, >>> Â Â which is why I am adding it. >> >> How about squashing this in? >> >> --8<-- >> Subject: Allow set_git_work_tree() to be called more than once >> >> Usually one of the setup functions is called once. However alias >> handling code needs to look ahead in $GIT_DIR/config for aliases. So >> set_git_work_tree() may be called twice: once when alias is searched, >> once when the actual command is run. >> >> Loosen the condition and let it through. We can stricten it back when >> alias handling is fixed. > > Sounds like sweeping something under rug. > > Without this hack, alias does not work, but if your plan is to rework this > part right when alias is fixed, then what's the point of loosening it? > Either way your alias is broken, no? > > Shouldn't it at least make sure that the function is giving a consistent > picture to the outside world between the time it was called for the first > time and the second time? ÂE.g. Âthe second time around new-work-tree must > be the same as whatever was given the first time, or something. Right. Will do that. -- Duy -- To unsubscribe from this list: send the line "unsubscribe git" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html