On Thu, Apr 16, 2009 at 1:26 PM, jeffhg58 <[email protected]> wrote:
> > I recently installed django 1.1 beta and I needed to run the django > test suite. So, under the tests directory I executed runtests.py. I > received some errors when trying to execute all the tests. I was > expected that all the tests would have passed. I am thinking it might > be with my setup. Here are the errors I encountered. > > http://dpaste.com/34433/ > First, I'm curious how you got tests for 1.1 beta? The tests weren't included in the tarfile until quite recently (post 1.1 beta, I believe, since my downloaded un-tarred copy of 1.1 does not have a tests directory). If you are getting tests out of svn you just need to be careful that you are running a version that matches the code you are using -- if you use more-recent tests against backlevel code you are liable to hit test failures resulting from bugfixes where the bugfix changeset has included a test that fails prior to the code fix. I don't believe that is what you are seeing here though. Second, the failure summary you point to lists 3 failure but I only see details of 2 failures listed above. Is that output complete? As for the failures: 1 - The Windows error on attempting to delete a file is something we've seen before, and fixed (though I'm not sure in this particular place). I cannot recreate the error on my Windows box with Python 2.6 and the beta 1.1 (nor current trunk) level code. What exact level of Python 2.6 are you running? -- the specifics of this error have been dependent on Python level in the past. Are you running anything on this machine (like a virus scanner) that may be opening these files created by the tests while the test itself is simultaneously trying to delete them? There was also at least one fix made after 1.1 beta that cleaned up some issues with temp files on Windows getting closed and deleted properly, so I'd be interested to know if you tried running with trunk level code if you still see this error. 2 - I've not seen the markup test error before but it looks like it may be textile-level dependent. I can recreate the problem with textile 2.1.3 (on Windows, where I previously had no textile installed) but not 2.0.10 (what I happen to have on Linux). The test may be overly sensitive to slight changes in what textile produces. I was going to say please open a ticket but upon searching I see this has already been reported: http://code.djangoproject.com/ticket/10843 On the general question as to whether all the test should pass -- in an ideal world, yes, but we are not quite there yet. There are OS and database backend issues that cause some known test failures. Linux/sqlite generally runs cleanly, Windows/sqlite may, but it is dependent on Python level (Python 2.5 on Windows has a particularly problematic sqlite level that it shipped with). I generally also see clean runs using MySQL/MyISAM (both Linux and Windows) but MySQL/InnoDB has many failures resulting from InnoDB's inability to do deferred constraint checking (there's a ticket open on that). Some levels of PostgreSQL show failures with the admin_views test (again, there's a ticket open on it). Most test failures I know of have either tickets open for them or doc notes ( http://docs.djangoproject.com/en/dev/ref/databases/#sqlite-3-3-6-or-newer-strongly-recommended) that cover them. Karen --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "Django users" group. To post to this group, send email to [email protected] To unsubscribe from this group, send email to [email protected] For more options, visit this group at http://groups.google.com/group/django-users?hl=en -~----------~----~----~----~------~----~------~--~---
