On Fri, Dec 5, 2008 at 1:20 AM, Daniel Shahaf <d.s_at_daniel.shahaf.name> wrote:
> Paul Burba wrote on Thu, 4 Dec 2008 at 14:16 -0500:
>> On Thu, Dec 4, 2008 at 8:56 AM, Daniel Shahaf <d.s_at_daniel.shahaf.name> wrote:
>> > Thanks for digging these up. After merging the actions.py part of r33974,
>> > the test at least runs, but fails on the directory case (which was added
>> > in r34053) due to missing UUID in the 'info' output (the file case
>> > passes). After I also merged r33775, r33782, r33786, and r33788, the
>> > whole test passed.
>> Ugh, really, I tried that and it still failed, that's when I went crazy :-P
> Yeah, it wasn't smooth. There were 3-4 conflicts to resolve on the way
> (see log of the branch).
>> > I'll look more closely later tonight.
>> A 1000 thank yous! Let us know what you find, I'll be happy to review
>> any nominations you make to get this working.
> I've backported all five changes to 1.5.x-r34050-followups branch now;
> could you look it if fixes merge_tests 89 for you? (It does for me.)
Yes, it passes.
> In the backports, I haven't yet looked into the details of the conflict
> that I fixed as part of r34570 (the 'depth' parameter differs). It's
> probably trivial (and I'll look into it before +1ing the backport), but
> for now that's the least-reviewed part of the backport branch.
Regarding the backported changes:
r34570: Always passing 'svn_depth_infinity' to svn_wc_ensure_adm3() is
what svn_wc_add2() now does on trunk, so that looks correct, +1.
r33974: Is there a reason we cannot backport the changes to
merge_tests.py and update_tests.py? This is the only choice I don't
Ran the [RA_LOCAL | RA_SVN] x [FSFS] tests on this branch, everything passes.
Nice use of the file target merges on the partial merges in
r34570-34571 BTW, it warms my cold heart to see that :-)
Received on 2008-12-08 10:41:40 CET