Re: Misleading 404 exception during unit tests.

2015-03-23 Thread Tim Graham
To have the decision revisited, you should raise it on django-developers. On Monday, March 23, 2015 at 10:39:57 AM UTC-4, Ben Scherrey wrote: > > Hey Tim, > >Appreciate the link. It looks like we're trying to have it go both ways > here. If we don't want to allow it into production then make

Re: Misleading 404 exception during unit tests.

2015-03-23 Thread Benjamin Scherrey
Hey Tim, Appreciate the link. It looks like we're trying to have it go both ways here. If we don't want to allow it into production then make it explicit and break it - just as before. If we want to recommend that it not go into production but then let users actually do it, don't silently break

Re: Misleading 404 exception during unit tests.

2015-03-23 Thread Tim Graham
Please see https://github.com/django/django/commit/4c6ffcf7 for the rationale for the change. I am open to ideas, but I'm concerned that including details about the reason for the 404 might cause information leakage in production which wouldn't be desirable. On Monday, March 23, 2015 at 10:06:2

Re: Misleading 404 exception during unit tests.

2015-03-23 Thread Benjamin Scherrey
I see now that the behaviour of serve() was changed for 1.7 and it now returns a silent misleading 404 result rather than the old ImproperlyConfigured exception which would have provided a clue as to the actual root cause. See: https://docs.djangoproject.com/en/1.7/ref/contrib/staticfiles/#django.