Phillip Wood [off-list ref] writes:
quoted
I suspect that it would be sufficient to make update_head() helper
function take the reflog action message as another parameter
instead to fix the above, but there may be other reasons why you
chose to do it this way---I cannot read it in your empty log
message, though.
Good point, sorry I should have added some explanation about that. I
went with using setenv() rather than passing a reflog message to
update_head() as it meant there were no changes needed on the sequencer
side as it already sets GIT_REFLOG_ACTION. As the sequencer already sets
GIT_REFLOG_ACTION, and git-commit does not fork any subprocesses I don't
Doesn't "git commit" run number of hooks? Is it just the current
code does not run any hooks (by chance) after these new setenv()
calls are made and we happen to be safe?