From: Junio C Hamano <gitster@pobox.com>
Date: Wed, 08 Apr 2020 21:51:11 GMT
Subject: Re: [PATCH] clean: explicitly `fflush` stdout
Message-ID: <xmqqr1wxoddc.fsf@gitster.c.googlers.com>
In-Reply-To: <20200408201454.GB2270445@coredump.intra.peff.net>

Jeff King <peff@peff.net> writes:

> This (and the patch) make sense to me. It might be worth factoring out a
> "read input from user" function and putting the flush there, but with
> only three affected call sites, I'm OK with it either way.
>
>>     This is yet another patch that was funneled through a Git for Windows
>>     PR. It has served us well for almost five years now, and it is beyond
>>     time that it find its final home in core Git.
>
> Agreed. I guess it didn't affect people on other platforms much because
> stdout to a terminal is generally line-buffered on POSIX systems. But
> it's better not to rely on that, plus it could create confusion if
> somebody tried to manipulate the interactive operation through a pipe
> (e.g., driving it from a GUI or something).

Hmph, I thought it was more common to do prompts etc. on the
standard error stream, which tends to make the buffering of the
output less of a problem, but apparently these prompts are given to
the standard output.  I am also OK to sprinkle fflush(stdout) but in
the longer term, it would probably be a good idea to introduce a
helper to "prompt then grab input" or "read user input" (if the
former, we'd be able to bring consistency into "which between the
standard output or the standard error does a prompt come?", and if
the latter we'd do fflush(NULL) before reading), especially if there
are many git subcommands that go interactive.

Thanks.