http://bugs.grommit.com/show_bug.cgi?id=474
------- Comment #2 from richlowe at richlowe.net 2008-04-15 22:13 PDT -------
Ethan can reproduce this locally, and, following what he said so I can I.
The problem is that when webrev tries to figure out the parent on its own, it
does this:
hg log -R $CWS --template '{node}\n' $CWS
That last $CWS is bogus, that's a file spec, so it lists changes to $CWS,
which is likely to turn out wrong in several instances, including this one,
as the tip revision (112) is a merge that's listed as changing no files of
its own, it does not touch $CWS, and as such, we don't even attempt to use
it, 111 is the newest revision that sees.
I believe removing the reference to $CWS is enough to fix it, but if we
look further up we see that we call hg-active even in this case (I think
the only case where we *don't* call hg-active is a wx file list). Given
we've already called it (either way), we could pull HG_PARENT out of it,
and save all this effort.
It's possibly worth while to mine hg-active for HG_PARENT even in the case
where we were passed a wx file list, and HG_PARENT was not set (I think that
would be uncommon, but covering bases here).
I'd obviously rather go the hg-active route, as it's less complex, and
shouldn't suffer from the "note: this will probably not work if you have
branches"
(http://src.opensolaris.org/source/xref/scm-migration/onnv-scm/usr/src/tools/scripts/webrev.sh#2121)
problem mentioned just above the referenced code.
--
Configure bugmail: http://bugs.grommit.com/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.
