SuspiciousOperations have been differentiated into subclasses, and
are now logged to a 'django.security.*' logger. SuspiciousOperations
that reach django.core.handlers.base.BaseHandler will now return a 400
instead of a 500.

Thanks to tiwoc for the report, and Carl Meyer and Donald Stufft
for review.

The test was always skipped on Python3 because string literals are unicode

Test expected that when given invalid utf-8, the backend should raise
a DatabaseError, but the Oracle backend raises a UnicodeDecodeError.

Thanks Joshua Fialkoff for the report.

Thanks Denis Martinez for the report and initial patch, and
Sergey Kolosov for bringing the patch up to date.

Updated core/admin/admindocs/comments translation catalogs.

Using two separate translation strings instead of gettext plural when
there is no reference to the number in the translated string. This
prevents some translations like Russian and Latvian to use the singular
form for 11 or 21.

Thanks Ram Rachum for the report and the initial patch, and Simon
Charette for the review.

 * LocaleMiddleware stores language into session if it is not present there.

Regression was introduced in 6a6bb168. Thanks Bas Peschier for the
report.

Updated link to jQuery Cookie plugin site

Updated link to active project for ODBC backend

It took me quite some time to find if and where the ODBC backend was maintained.
I found (on djangoproject.com):
http://code.google.com/p/django-pyodbc/ (last commit about 3 years ago)
then:
https://github.com/avidal/django-pyodbc avidal fork.
then:
https://github.com/aurorasoftware/django-pyodbc/ aurorasoftware version which has avidal improvements merged.

Avidals version now links to https://github.com/aurorasoftware/django-pyodbc/ which is also the version installed through PIP.

Thanks to av@rdf.ru and flarno11@yahoo.de for the report.

Allows a `GenericForeignKey` to reference proxy models. The default
for `for_concrete_model` is `True` to keep backwards compatibility.

Also added the analog `for_concrete_model` kwarg to
`generic_inlineformset_factory` to provide an API at the form level.

Only make assertions that actually test the functionality of the
change and not unrelated pieces.

Thanks uberj@ for the report and lolek09 for the patch.

Refs #20474.

Refs #20474.

Thanks Katya for the suggestion.

Thanks diek@ for the suggestion.

Thanks jacob for the suggestion and ryankask
for the patch.

Don't try to be smart about building a good-looking help string
because it evaluates translations too early, simply use the same old
strategy as before. Thanks Donald Stufft for the report.

Also, actually fix the case reported by the OP by special-casing
CheckboxSelectMultiple.

Added tests.

Refs #9321.

Fixed comments for translators example in the i18n documentation

The documentation regarding adding a comment for translators was confusing
as the example contained mismatching tags.

Fixed #11398 - Added a pre_syncdb signal