On Fri, May 20, 2011 at 4:21 PM, Philip Martin <philip.mar...@wandisco.com> wrote: > Hyrum K Wright <hy...@hyrumwright.org> writes: > >> This is already nominated elsewhere in the file. >> >> (Hint: look in the 'Vetoed' section. :) ) > > Hmm, > > -1: hwright (I'm not sure we need this: the merge conflicts, and the > assert isn't even in 1.6.x. I'm happy to remove this veto > if a proper backport branch is provided.) > > The assert is definitely in 1.6 as it causes svnsync_tests.py 29 to fail > both on the buildbot and on my machine: > > $ nice make check FS_TYPE=bdb TESTS=subversion/tests/cmdline/svnsync_tests.py > Running all tests in svnsync_tests.py [1/1]...FAILURE > At least one test was SKIPPED, checking > /home/pm/sw/subversion/obj-1.6/tests.log > SKIP: svnsync_tests.py 14: verify that unreadable content is not synced > SKIP: svnsync_tests.py 15: verify that copies from unreadable dirs work > SKIP: svnsync_tests.py 16: verify copies with mods from unreadable dirs > SKIP: svnsync_tests.py 17: verify copies with mods from unreadable dirs +copy > At least one test FAILED, checking /home/pm/sw/subversion/obj-1.6/tests.log > FAIL: svnsync_tests.py 29: descending into replaced dir looks in src > Summary of test results: > 4 tests SKIPPED > 1 tests FAILED > make: *** [check] Error 1 > > It merges cleanly: > > $ svn merge -cr1036429 ^/subversion/trunk ../src-1.6 > --- Merging r1036429 into '../src-1.6': > U ../src-1.6/subversion/libsvn_repos/replay.c > > > I guess it was not clear that this is part of the issue 3641 fix, > i.e. we need it now that the issue 3641 branch has been merged.
That's probably it. I hadn't seen the failures you mention since the branch got merged while I was in the air. I tested the merge, and it works fine, so I removed my veto and moved this to Approved. -Hyrum
