From: Junio C Hamano <junkio@cox.net>
Date: Sun, 08 Jan 2006 02:22:54 GMT
Subject: Re: [PATCH 1/7] Add a minimal test for git-cherry
Message-ID: <7vy81r5ugx.fsf@assigned-by-dhcp.cox.net>
In-Reply-To: <20060108003815.GI1113@nowhere.earth>

Yann Dirson <ydirson@altern.org> writes:

> This test checks that git-cherry finds the expected number of patches
> in two simple cases.
>
> Signed-off-by: Yann Dirson <ydirson@altern.org>
>
> +test_expect_success \
> +    'cherry-pick one of the 2 patches, and check cherry recognized one and only one as new' \
> +    'git-cherry-pick my-topic-branch^0 &&
> +     echo $(git-cherry master my-topic-branch) &&
> +     expr "$(echo $(git-cherry master my-topic-branch) )" : "+ [^ ]* - .*"
> +'

I wonder if "+ followed by -" order is stable here.  If you make
commits in the my-topic-branch fast enough, they would get the
same timestamp and I suspect the order from rev-list would be
left to chance.