On Thu, Jul 4, 2013 at 2:46 PM, Orgad Shaneh <org...@gmail.com> wrote:
> On Thu, Jul 4, 2013 at 3:42 PM, Antoine Pelisse <apeli...@gmail.com> wrote:
>>>> Your problem is that your hook script is not checking $2 so it is
>>>> overwriting the message even when you do not want to do so.
>>> No, it isn't. Not by git-gui at least. Check /tmp/hook.log with the
>>> hook I provided...
>> So what you mean is that the hook is not executed with the correct
>> Could you please provide the output of the /tmp/hook.log file (I can't
>> reproduce right now) ?
> It only runs for "New commit" (possibly with merge or squash as
> arguments). In my case the log is:
> Not running the hook for amend is another problem.
> What I referred to was that handling the hook's result is done without
> checking if the state has changed while it was running, like Fredrik
> has already pointed out.
Yep, I've had a quick look at that, and clearly the problem seems to
come from git-gui.
I have absolutely no knowledge of TCL so I won't be able to help.
Indeed, from what I've seen it looks like prepare-commit-message hook
is not called in amend case.
I cc'ed Pat Thoyts as he's the maintainer of git-gui.
Sorry for not being able to help any further on this one.
To unsubscribe from this list: send the line "unsubscribe git" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html