[PATCH] parsers: it should check the other interesting before decreasing ninteresting (issue3984)

[Siddharth Agarwal]

On 07/25/2013 05:38 AM, Chingis Dugarzhapov wrote:
> Hi Elson,
>
> The result of hg log -r "ancestors(10,9)" for 2.2.2 and 2.6.3+patch is 
> different for me:
>
> 2.2.2:
>
> # hg log -r "ancestor(10,9)"
> changeset:   4:f303d6efbf44
> branch:      foo
> parent:      2:d0f226159ee1
> user:        Tavis Elliott <telliott at apptio.com 
> <mailto:telliott at apptio.com>>
> date:        Wed Jul 17 09:11:23 2013 -0700
> summary:     Added e, modified c
>
> 2.6.3+patch:
>
> # hg log -r "ancestor(10,9)"
> changeset:   6:67c8c7ad41bf
> branch:      foo
> parent:      2:d0f226159ee1
> user:        Tavis Elliott <telliott at apptio.com 
> <mailto:telliott at apptio.com>>
> date:        Wed Jul 17 09:07:13 2013 -0700
> summary:     added d, changed c

Actually, this looks OK -- 2.2.2's output is wrong because it picks the 
lowest rev number and not the lowest SHA1, as explained in my other 
email. Note that 6 has a lower SHA1 than 4.