Le 14/07/2010 11:28, Baiju M a écrit :

On Wed, Jul 14, 2010 at 2:57 PM, Baiju M<mbaiju@zeomega.com> wrote:

...

On Sun, Jul 11, 2010 at 5:49 PM, Christophe Combelles<ccomb@free.fr> wrote:

...

Le 11/07/2010 13:59, Zope Tests Summarizer a écrit :

...

Subject: FAILED : Bluebream / Python2.4.6 32bit linux From: ccomb at free.fr Date: Sat Jul 10 22:05:41 EDT 2010 URL: http://mail.zope.org/pipermail/zope-tests/2010-July/016427.html

Subject: FAILED : Bluebream / Python2.5.2 32bit linux From: ccomb at free.fr Date: Sat Jul 10 22:07:33 EDT 2010 URL: http://mail.zope.org/pipermail/zope-tests/2010-July/016428.html

Subject: FAILED : Bluebream / Python2.6.4 32bit linux From: ccomb at free.fr Date: Sat Jul 10 22:07:34 EDT 2010 URL: http://mail.zope.org/pipermail/zope-tests/2010-July/016429.html

This is a failure in z3c.testsetup. Can someone have a look at it? It also fails with 0.7.0.

FYI: In trunk, if I pin zc.recipe.testrunner to 1.2.0, this failure is going away.

I mean the trunk of "z3c.testsetup"

Ok thanks, I've also checked that. Can someone release a new version of z3c.testsetup (say 0.8)? Or add me as owner so I can do it? (pypi: ccomb) regards, Christophe

Regards, Baiju M

Le 24/07/2010 03:51, Uli Fouquet a écrit :

Hi there,

Christophe Combelles wrote:

...

Le 14/07/2010 11:28, Baiju M a écrit :

...

On Wed, Jul 14, 2010 at 2:57 PM, Baiju M<mbaiju@zeomega.com> wrote:

...

On Sun, Jul 11, 2010 at 5:49 PM, Christophe Combelles<ccomb@free.fr> wrote:

...

Le 11/07/2010 13:59, Zope Tests Summarizer a écrit :

...

Subject: FAILED : Bluebream / Python2.4.6 32bit linux From: ccomb at free.fr Date: Sat Jul 10 22:05:41 EDT 2010 URL: http://mail.zope.org/pipermail/zope-tests/2010-July/016427.html

Subject: FAILED : Bluebream / Python2.5.2 32bit linux From: ccomb at free.fr Date: Sat Jul 10 22:07:33 EDT 2010 URL: http://mail.zope.org/pipermail/zope-tests/2010-July/016428.html

Subject: FAILED : Bluebream / Python2.6.4 32bit linux From: ccomb at free.fr Date: Sat Jul 10 22:07:34 EDT 2010 URL: http://mail.zope.org/pipermail/zope-tests/2010-July/016429.html

This is a failure in z3c.testsetup. Can someone have a look at it? It also fails with 0.7.0.

FYI: In trunk, if I pin zc.recipe.testrunner to 1.2.0, this failure is going away.

I mean the trunk of "z3c.testsetup"

Ok thanks, I've also checked that.

Can someone release a new version of z3c.testsetup (say 0.8)? Or add me as owner so I can do it? (pypi: ccomb)

Sorry for reacting so slowly and thanks for the hint!

Just released z3c.testsetup 0.8 and tests seem to work again at least on my test system with Linux 32-bit and with different ZTKs. Locally also bluebream seems to be happy with the new piece.

Instead of pinning zc.recipe.testrunner I took the liberty to fix the tests. Hope that works for other constellations as well.

thanks!! There is only one remaining error on Python 2.4 now. It seems related to an 'encoding' argument added to DocFileSuite in python 2.5, and used in zope.app.testing Christophe

I also added you as owner of the package on PyPI, to avoid such a blocking situation next time.

Best regards,

Hi Christophe,

Christophe Combelles wrote:

[snip: z3c.testsetup test errors on bluebream]

...

thanks!! There is only one remaining error on Python 2.4 now. It seems related to an 'encoding' argument added to DocFileSuite in python 2.5, and used in zope.app.testing

Oh, missed that. This is more tricky but I'll have a look into it. Thanks for the hint (and the buildbots revealing such nasty problems)!

Thanks for releasing 0.8.1. It seemed to fix it! However I have a weird problem: builders were all green, then I suddenly had repeated failures with 0.8.1 on all the builders, then I've tried to launch the tests by myself directly in the builder, and tests passed, without changing anything... Then if I relaunch the builder alone, tests pass: http://buildbot.afpy.org/bluebream/waterfall Is there a possibility that tests fail when several builders run at the same time? File "/home/ccomb/buildout-eggs/z3c.testsetup-0.8.1-py2.6.egg/z3c/testsetup/README.txt", line 110, in README.txt Failed example: testrunner.run(defaults) Differences (ndiff with -expected +actual): - Running z3c...layer.DefaultZCMLLayer [...ftesting.zcml] tests: - Set up z3c...layer.DefaultZCMLLayer [...ftesting.zcml] in N.NNN seconds. + Running z3c.testsetup.functional.layer.DefaultZCMLLayer [/home/ccomb/buildout-eggs/z3c.testsetup-0.8.1-py2.6.egg/z3c/testsetup/tests/othercave/ftesting.zcml] tests: + Set up z3c.testsetup.functional.layer.DefaultZCMLLayer [/home/ccomb/buildout-eggs/z3c.testsetup-0.8.1-py2.6.egg/z3c/testsetup/tests/othercave/ftesting.zcml] in N.NNN seconds. Ran 3 tests with 0 failures and 0 errors in N.NNN seconds. - Running z3c...layer.DefaultZCMLLayer [...ftesting2.zcml] tests: - Tear down z3c...layer.DefaultZCMLLayer [...ftesting.zcml] ... not supported + Running z3c.testsetup.functional.layer.DefaultZCMLLayer [/home/ccomb/buildout-eggs/z3c.testsetup-0.8.1-py2.6.egg/z3c/testsetup/tests/othercave/ftesting2.zcml] tests: + Tear down z3c.testsetup.functional.layer.DefaultZCMLLayer [/home/ccomb/buildout-eggs/z3c.testsetup-0.8.1-py2.6.egg/z3c/testsetup/tests/othercave/ftesting.zcml] ... not supported Running in a subprocess. - Set up z3c...layer.DefaultZCMLLayer [...ftesting2.zcml] in N.NNN seconds. + Set up z3c.testsetup.functional.layer.DefaultZCMLLayer [/home/ccomb/buildout-eggs/z3c.testsetup-0.8.1-py2.6.egg/z3c/testsetup/tests/othercave/ftesting2.zcml] in N.NNN seconds. Ran 1 tests with 0 failures and 0 errors in N.NNN seconds. - Tear down z3c...layer.DefaultZCMLLayer [...ftesting2.zcml] ... not supported + Tear down z3c.testsetup.functional.layer.DefaultZCMLLayer [/home/ccomb/buildout-eggs/z3c.testsetup-0.8.1-py2.6.egg/z3c/testsetup/tests/othercave/ftesting2.zcml] ... not supported Running z3c.testsetup.tests.othercave.testing.FunctionalLayer1 tests: Running in a subprocess. Set up z3c.testsetup.tests.othercave.testing.FunctionalLayer1 in N.NNN seconds. Ran 1 tests with 0 failures and 0 errors in N.NNN seconds. Tear down z3c.testsetup.tests.othercave.testing.FunctionalLayer1 in N.NNN seconds. Running z3c.testsetup.tests.othercave.testing.UnitLayer2 tests: Running in a subprocess. Set up z3c.testsetup.tests.othercave.testing.UnitLayer1 in N.NNN seconds. Set up z3c.testsetup.tests.othercave.testing.UnitLayer2 in N.NNN seconds. Running testSetUp of UnitLayer1 Running testSetUp of UnitLayer2 Running testTearDown of UnitLayer2 Running testTearDown of UnitLayer1 Ran 1 tests with 0 failures and 0 errors in N.NNN seconds. Tear down z3c.testsetup.tests.othercave.testing.UnitLayer2 in N.NNN seconds. Tear down z3c.testsetup.tests.othercave.testing.UnitLayer1 in N.NNN seconds. - Running zope...testrunner.layer.UnitTests tests: ? ^ + Running zope.testing.testrunner.layer.UnitTests tests: ? ^^^^^^^ Running in a subprocess. - Set up zope...testrunner.layer.UnitTests in N.NNN seconds. ? ^ + Set up zope.testing.testrunner.layer.UnitTests in N.NNN seconds. ? ^^^^^^^ - Custom setUp for <DocTest doctest05.txt from ...doctest05.txt:0 (2 examples)> - Custom tearDown for <DocTest doctest05.txt from ...doctest05.txt:0 (2 examples)> + Custom setUp for <DocTest doctest05.txt from /home/ccomb/buildout-eggs/z3c.testsetup-0.8.1-py2.6.egg/z3c/testsetup/tests/othercave/doctest05.txt:0 (2 examples)> + Custom tearDown for <DocTest doctest05.txt from /home/ccomb/buildout-eggs/z3c.testsetup-0.8.1-py2.6.egg/z3c/testsetup/tests/othercave/doctest05.txt:0 (2 examples)> Ran 4 tests with 0 failures and 0 errors in N.NNN seconds. - Tear down zope...testrunner.layer.UnitTests in N.NNN seconds. ? ^ + Tear down zope.testing.testrunner.layer.UnitTests in N.NNN seconds. ? ^^^^^^^ - Total: 10 tests, 0 failures, 0 errors in N.NNN seconds. + Total: 10 tests, 0 failures, 0 errors in 1 minutes N.NNN seconds. ? ++++++++++ False

Best regards,

Hi Christophe, Christophe Combelles wrote:

Thanks for releasing 0.8.1. It seemed to fix it!

However I have a weird problem: builders were all green, then I suddenly had repeated failures with 0.8.1 on all the builders, then I've tried to launch the tests by myself directly in the builder, and tests passed, without changing anything... Then if I relaunch the builder alone, tests pass: http://buildbot.afpy.org/bluebream/waterfall

Is there a possibility that tests fail when several builders run at the same time?

Yes, and I think for a strange reason: they took to much time (probably because they share system resources and simply run slower each). The problem here was, that in the test testrunners are run and the regular expressions matching their output did expect output of kind "Total: ... 0 errors in N.NNN seconds" and not: "Total: ... 0 errors in 1 minutes N.NNN seconds" But that's what happened in buildbot runs: they took more than one minute to run some subprocesses. On a real slow and old machine you might be able to reproduce the error :-) I think there is no need for special configuration or similar. Instead, I fixed the regular expressions and fixed also some other things, so that the next release should be okay (also with Python 2.7). The changes are alreadey checked in, just had no time to do the actual release. Please expect 0.8.2 this evening. May I update the bluebream versions.cfg then as well? Until now Baiju was friendly enough to do that (thanks for that!). Sorry for the trouble! Hope, there will be less problems then. Best regards, -- Uli