From: Christian Couder <chriscool@tuxfamily.org>
Date: Fri, 28 Aug 2009 05:31:12 GMT
Subject: Re: [PATCH] git-bisect: call the found commit "*the* first bad commit"
Message-ID: <200908280731.12827.chriscool@tuxfamily.org>
In-Reply-To: <20090826173850.6117@nanako3.lavabit.com>

On Wednesday 26 August 2009, Nanako Shiraishi wrote:
> .. as we learned in the school ;-)
>
> Signed-off-by: Nanako Shiraishi <nanako3@lavabit.com>
> ---
>  bisect.c                    |    2 +-
>  git-bisect.sh               |    2 +-
>  t/t6030-bisect-porcelain.sh |   18 +++++++++---------
>  3 files changed, 11 insertions(+), 11 deletions(-)
>
> diff --git a/bisect.c b/bisect.c
> index 7f20acb..dc18db8 100644
> --- a/bisect.c
> +++ b/bisect.c
> @@ -991,7 +991,7 @@ int bisect_next_all(const char *prefix)
>
>  	if (!hashcmp(bisect_rev, current_bad_sha1)) {
>  		exit_if_skipped_commits(tried, current_bad_sha1);
> -		printf("%s is first bad commit\n", bisect_rev_hex);
> +		printf("%s is the first bad commit\n", bisect_rev_hex);
>  		show_diff_tree(prefix, revs.commits->item);
>  		/* This means the bisection process succeeded. */
>  		exit(10);

Thanks, but I wonder if this could give the false impression that there can 
only be one "first bad commit". Because it's possible that a bug appears in 
one commit say A, then get fixed in another one say B, and eventually 
reappears in a third one say C. So if the bisection range contains all 
these commits, the result from bisecting could be A or C.

Best regards,
Christian.