From: Tay Ray Chuan <rctay89@gmail.com>
Date: Thu, 11 Feb 2010 22:54:47 GMT
Subject: Re: [PATCH 3/4] git-push: send "To <remoteurl>" messages to the standard output in --porcelain mode
Message-ID: <be6fef0d1002111454w10ad965ej72699d1e185d2a9f@mail.gmail.com>
In-Reply-To: <45357037fd8f928759c9e9064736c62e793b84be.1265694627.git.larry@elder-gods.org>

Hi,

On Tue, Feb 9, 2010 at 1:54 PM, Larry D'Anna <larry@elder-gods.org> wrote:
> diff --git a/transport.c b/transport.c
> index 3846aac..fb653c6 100644
> --- a/transport.c
> +++ b/transport.c
> @@ -675,7 +675,7 @@ static void print_ok_ref_status(struct ref *ref, int porcelain)
>  static int print_one_push_status(struct ref *ref, const char *dest, int count, int porcelain)
>  {
>        if (!count)
> -               fprintf(stderr, "To %s\n", dest);
> +               fprintf(porcelain ? stdout : stderr, "To %s\n", dest);
>
>        switch(ref->status) {
>        case REF_STATUS_NONE:

just to check - are debug and user-specific messages (like advice and
info) to be sent, by convention/design/force-of-habit, to stderr? Are
those the "cruft" that has been referred to in other threads?

If that's the case, then perhaps the "To: <destination>" lines should
be sent to both stdout and stderr - stdout, for porcelain scripts as a
"header", and stderr, to help the user make sense of any errors that
occur ("oh, these errors were triggered when pushing to so-and-so
remote").

-- 
Cheers,
Ray Chuan