From: Junio C Hamano <gitster@pobox.com>
Date: Thu, 20 Jun 2013 18:11:00 GMT
Subject: Re: [PATCH] status: display the SHA1 of the commit being currently processed
Message-ID: <7vzjukveiz.fsf@alter.siamese.dyndns.org>
In-Reply-To: <51C2B8DA.6040400@viscovery.net>

Johannes Sixt <j.sixt@viscovery.net> writes:

> Am 6/20/2013 9:56, schrieb Peter Krefting:
>> Junio C Hamano:
>> 
>>> But my understanding is that the reordering using printf() is the
>>> mechanism we suggest l10n folks to use when the order of parameters
>>> given to printf does not match the preferred word order in the message
>>> in their language.
>> 
>> It's documented in the gettext manual, and seems to be used in the
>> zh_CN.po to change the word order in quite a few places.
>
> It is fine to use %n$ in translated strings as long as gettext is enabled
> only on systems that have a sufficiently capable printf and these formats
> are not used in the source code.
>
> But you can't have this string:
>
>   "Splitting a commit while rebasing branch '%2$s' on '%3$s'."
>
> neither in the template nor in the translation, because the numbers must
> begin at 1 (and must be used without gaps).

Did any message we saw in the patch (and the discussion to possibly
improve it) need to have such a format string, or are you pointing
out a common gotcha we may want to warn translators about in
po/README?