Re: git cherry-pick conflict error message is deceptive when cherry-picking multiple commits

On Wed, Jul 27, 2016 at 11:03 AM, Johannes Schindelin
[off-list ref] wrote:

Hi Stephen,

On Wed, 27 Jul 2016, Stephen Morton wrote:

quoted

Here is my patch then. (Personally, I would add some capitalization and
punctuation, but I didn't see much of that in the existing code.) I'm
not a regular pull-requester, do I do that, or can somebody else handle
that for me?

The process of the patch submission is described in
Documentation/SubmittingPatches (yes, it is a bit involved, and it is
slightly easier when you use http://submitgit.herokuapp.com/, but please
note that this process has served us well over one decade).

Please also note that top-posting is highly discouraged on this list:

A: Because it messes up the order in which people normally read text.

quoted

Q: Why is top-posting such a bad thing?

quoted

A: Top-posting.

quoted

Q: What is the most annoying thing in e-mail?

Now to your patch:

quoted

diff --git a/sequencer.c b/sequencer.c
index cdfac82..ce06876 100644
--- a/sequencer.c
+++ b/sequencer.c

@@ -176,7 +176,8 @@ static void print_advice(int show_hint, struct

replay_opts *opts)
                else
                        advise(_("after resolving the conflicts, mark
the corrected paths\n"
                                 "with 'git add <paths>' or 'git rm <paths>'\n"
-                                "and commit the result with 'git commit'"));
+                                "then continue the %s with 'git %s
--continue'\n"
+                                "or cancel the %s operation with 'git
%s --abort'" ),  action_name(opts), action_name(opts),
action_name(opts), action_name(opts));

That is an awful lot of repetition right there, with an added
inconsistency that the action is referred to by its name alone in the
"--continue" case, but with "operation" added in the "--abort" case.

And additionally, in the most common case (one commit to cherry-pick), the
advice now suggests a more complicated operation than necessary: a simply
`git commit` would be enough, then.

Can't we have a test whether this is the last of the commits to be
cherry-picked, and if so, have the simpler advice again?

Ciao,
Johannes

Ok, knowing that I'm not on the last element of the sequencer is
beyond my git code knowledge. I see that in do_pick_commit() , we do
not have a copy of the todo_list. I would assume that would be
necessary, but I'm not certain. I could
file_exists(git_path_seq_dir()). This works to determine if one or
many commits are being cherry-picked / reverted, although it will
return true even on the last of n cherry-picks. I think that is still
reasonable.

I was trying to just take the same text as 'git status' already
displays. It could indeed be made more concise.

Happy to use the submission process, I just didn't know it. Thanks for
letting me know.

(Yup, sorry about the top-posting. I just wan't careful.)

Stephen