Re: [PATCH] rebase -i: remove CHERRY_PICK_HEAD when cherry-pick failed

[Andrew Wong <andrew.w@sohovfx.com>]

On 04/04/2012 02:11 PM, Jonathan Nieder wrote:
> Do you mean we should get rid of CHERRY_PICK_HELP?
>   
No. I thought you meant that if "cherry-pick" runs into an error, it
should not leave behind a state (i.e. CHERRY_PICK_HEAD) when
CHERRY_PICK_HELP is defined. And I was arguing that defining
CHERRY_PICK_HELP shouldn't affect the behavior of "cherry-pick" at all.
So it shouldn't be trying to remove the state in the first place. The
cleanup responsibility should fall into caller of "cherry-pick". i.e.
"rebase -i"

Though I now think that my original patch description could be improved
to better reflect that.

And you also mention earlier that the patch is more of a symptom relief,
and that
> a more appropriate long-term fix would involve "git
> cherry-pick" noticing when a patch has resolved to nothing instead of
> leaving it to "git commit" to detect that.
And I was arguing that "cherry-pick" doesn't have to detect scenarios
where "commit" could fail. Since there could be other scenarios where
"commit" could fail and "cherry-pick" is already handling "commit"
failing, I think there's no need for "cherry-pick" to handle an empty
commit specifically.

So if the list or Junio thinks that the patch is the right thing to do,
I should improve on the patch description before we queue it.