From: Jeff King <peff@peff.net>
Date: Sun, 03 Feb 2008 07:13:18 GMT
Subject: Re: [RFH] revision limiting sometimes ignored
Message-ID: <20080203071318.GA13849@coredump.intra.peff.net>
In-Reply-To: <7vr6fuzgq1.fsf@gitster.siamese.dyndns.org>

On Sat, Feb 02, 2008 at 10:39:18PM -0800, Junio C Hamano wrote:

> It is not so simple, it appears.  If I add for reproducibility
> "test_tick" like this:
> 
>         commit () {
>                 test_tick &&
>                 echo $1 >file &&
>                 git commit -a -m $1 &&
>                 git tag $1
>         }

Ah. I think what is happening is something like this:

  - when we add 'four' as uninteresting, we mark its parents as
    uninteresting in handle_commit
  - we don't recursively follow all of its parents because we haven't
    parsed them yet
  - when we get to limit_list, we call mark_parents_uninteresting again.
    But we have already marked four^ as uninteresting, and therefore we
    do not recurse in marking
  - we add the parents to the list, but they are not interesting, and
    therefore we quit

The reason it works with test_tick is that it changes the order we deal
with the commits in limit_list. We deal with 'one' _after_ dealing with
the uninteresting parents, so we never bail with
everybody_uninteresting.

> +test_expect_failure 'one is ancestor of others and should not be shown' '
> +
> +	git rev-list one --not four >result &&
> +	>expect &&
> +	diff -u expect result 
> +
> +'

Hooray, test_expect_failure is used properly. But you are still
missing a test_done. :)

-Peff