Re: [PATCH v3] git-merge: rewrite already up to date message

From: Josh Soref <jsoref@gmail.com>
Date: May 2, 2021, 01:51 UTC
Message-ID: <CACZqfqCXrhUtsC3TLYaw6u7D0VohsBcz10aVxpe-1u8gg77qxQ@mail.gmail.com>
In-Reply-To: <xmqqk0ol7cka.fsf@gitster.g>

"Josh Soref via GitGitGadget" <gitgitgadget@gmail.com> writes:

> From: Josh Soref <jsoref@gmail.com>
> Co-authored-by: Eric Sunshine <sunshine@sunshineco.com>
> Signed-off-by: Josh Soref <jsoref@gmail.com>

Junio C Hamano <gitster@pobox.com> writes:

> I am not sure why this is Co-au, and not the more usual "Helped-by".

If you look at the thread, you'll see that the code in question was written by Eric [1]. The only change from it was the addition of `void` to the function prototype by me.

That said, it's my first foray into patches for git. I didn't use helped-by because gitgitgadget didn't tell me to [2].

I have no opinion on the details. At this point, it's likely that a second commit would include a helped-by referencing you. Maybe. I'm not sure of the semantics of helped-by [2].

Junio C Hamano <gitster@pobox.com> wrote:

Show 12 quoted lines

> Actually, not so fast.  The end-users do not care really where the
> message originates.
>
> $ git grep -e 'Already up[- ]to' \*.c
> maint:builtin/merge.c:          finish_up_to_date(_("Already up to date."));
> maint:builtin/merge.c:                  finish_up_to_date(_("Already up to date. Yeeah!"));
> maint:merge-ort-wrappers.c:             printf(_("Already up to date!"));
> maint:merge-recursive.c:                output(opt, 0, _("Already up to date!"));
> maint:notes-merge.c:                    printf("Already up to date!\n");
>
> It probably makes sense to replace the exclamation point with a full
> stop for others, no?

Maybe. I'm not sure what they mean.

I'm fully capable of wearing a grammar / UX hat and rewriting the entire UX for a project if you invite me to do so.

I generally try not to do that when I initially approach a project, I prefer to get more comfortable w/ it and let it get more comfortable w/ me before I make significant change proposals.

(As an aside, I did start looking into what these messages meant / what to change them to, but other things have come up, and I've decided that I should at least respond to this message instead of just appearing to disappear.)

Show 10 quoted lines

> Also, I didn't notice when reading the patch submission earlier, but
> what does
>
> >> (All localizations of the previous strings are broken.)
>
> mean, exactly?  Do you mean to say
>
>     Because this changes some messages, the old messages that were
>     already translated will no longer be used, and these new
>     messages need to be translated anew.

Yes, this is what I meant. It's probably technically obvious to some people, but since I'm invited to describe the effects of my change, it seemed worth noting. Anyone cutting a release with this commit but not updating the translations would be bleeding en-US into the translations where before they would have had translated content. (Whether that translated content was any good given the fact that it was pasted together is a different story, but...)

[1] https://lore.kernel.org/git/CAPig+cT=xeLn9KNHz7hiNWo0QTfc1zZ1X-czJ4n503RRhBA0XQ@mail.gmail.com/ [2] https://github.com/gitgitgadget/gitgitgadget/issues/543