Re: [PATCH v6 0/7] New proc-receive hook for centralized workflow

From: Junio C Hamano <gitster@pobox.com>
Date: Apr 2, 2020, 18:26 UTC
Message-ID: <xmqqftdl3fss.fsf@gitster.c.googlers.com>
In-Reply-To: <20200402163520.1338-1-worldhello.net@gmail.com>

Jiang Xin <worldhello.net@gmail.com> writes:

Show 9 quoted lines

> From: Jiang Xin <zhiyou.jx@alibaba-inc.com>
>
> ## Changes since v5
>
> + Fixed code review issues on v5.
>
>   * s/cannot to/cannot/
>
>   * s/FLEX_ALLOC_MEM/FLEX_ALLOC_STR/

OK

>   * Add flag `C_LOCALE_OUTPUT` on test cases which may break
>     gettext-poison test.

We prefer to avoid the C_LOCALE_OUTPUT prerequiste when we can use test_i18n<stuff> to make the poison test not care. If it is too hard to arrange in these tests, it is OK to omit, though.

>   * Note: `rp_errors()` in "receive-pack" is executed on the
>     server side, I think leave them unmarked for i18n is better.

Hmph, I vaguely recall saying something similar and then got a disagreement from somebody. Perhaps Jonathan (cc'ed)?

Thanks.