[svn.haxx.se] · SVN Dev · SVN Users · SVN Org · TSVN Dev · TSVN Users · Subclipse Dev · Subclipse Users · this month's index

Re: 1.6.x and r38000

From: Julian Foad <julianfoad_at_btopenworld.com>
Date: Wed, 30 Sep 2009 00:50:22 +0100

On Tue, 2009-09-29, Stephen Butler wrote:
> The current state of the r38000 group looks good to me, except for a
> couple of test failures over HTTP. I'm running on OS X 10.6 with
> apr 1.3.8, httpd 2.2.12, neon 29.0, and serf 0.3.0. The failures
> are:
> merge_tests.py 133
> Over HTTP, there's a redundant notification of a tree-conflicted dir
> (A/B/E: "local edit, incoming delete upon merge"). This test is new.
> It's had problems before, and is currently skipped over serf.
> Skip(merge_replace_causes_tree_conflict,
> svntest.main.is_ra_type_dav_serf),
> If I get rid of the Skip() I get the same results over neon and serf.
> Does this test fail over neon for anyone else?

This test passes on the Neon build-bot (OSX 10.4, apr-0.9.x,
httpd-2.0.55, neon-0.25.5).

I note that the consequence of this test failing in the way it does with
Serf looks trivial: just an extra "Skipped" notification along with the
tree conflict notification. Not that "looks trivial" is any guarantee of
correct operation.

> switch_tests.py 19
> This test changes a prop of a dir that's been deleted in the repo,
> then commits. The commit should fail with "out of date". Over HTTP
> it succeeds.

On the Serf build-bot the commit succeeds so the test fails. On the Neon
build-bot the commit fails so the test passes.

> Looking at the httpd access_log, I see the old "PROPPATCH that
> succeeds" behavior mentioned a while back on the dev-list.
> http://svn.haxx.se/dev/archive-2009-05/0338.shtml
> But it looks like r37745 was already merged to 1.6.x. This test fails
> on ^/branches/1.6.x, too. IMHO it shouldn't delay merging the r38000
> group.


You can see all the build-bot results for this branch here:

- Julian

Received on 2009-09-30 01:50:44 CEST

This is an archived mail posted to the Subversion Dev mailing list.

This site is subject to the Apache Privacy Policy and the Apache Public Forum Archive Policy.