Also systematically added non-ascii passwords in hashers test suite.
Thanks Vaal for the report.

Thanks aliva for the report and claudep for the feedback.

This regression was introduced by the unicode_literals patch. The WSGI
spec mandates that environ contains native strings.

Thanks Tim Graham for the review and suggestions.

Thanks Mike O'Connor for the report.

Thanks Anton Baklanov for the report and the patch.

Caused by regression fix for #19500.

The added promotion logic is based on promoting any joins used in only
some of the childs of an OR clause unless the join existed before the
OR clause addition.

The ORM didn't reuse joins for direct foreign key traversals when using
chained filters. For example:
    qs.filter(fk__somefield=1).filter(fk__somefield=2))
produced two joins.

As a bonus, reverse onetoone filters can now reuse joins correctly

The regression was caused by the join() method refactor in commit
68847135

Thanks for Simon Charette for spotting some issues with the first draft
of the patch.

This is necessary because get_model() checks are case insensitive, and if the swapable check isn't, the
swappable logic gets tied up in knots with models that are partially swapped out.

Thanks to chris@cogdon.org for the report and extensive analysis, and Preston for his work on the draft patch.

Thanks to Patryk Zawadzki for the patch.

Thanks orblivion for the suggestion.

Thanks startup.canada for the suggestion.

Thanks jcspray for the suggestion.

Fixed #19430 -- Added 'license' value to setup.py

Thanks Dylan Verheul for the report and Anssi Kääriäinen for the
review.

Thanks to Simon Charette for report.

Refs #9589.

Cleaned up the implementation of base convert_values() a little, and
made sure it accepts None as a value for numeric fields.

There are no tests attached. The reason is that not all of the
convert_values() accept None as a value for numeric fields (for example
sqlite3.convert_values()).

The reason the base convert_values() needs to accept None is that this
situation might arise in custom compilers for 3rd party backends. It
is easy to keep the convert_values() working, so lets do that.

Thanks colinnkeenan for the suggestion.

Thanks Valentin Lorentz for the report and the suggested fix.

The OmniGraffle source file is included for future use.

Thanks Tim Graham for reporting this issue and testing the fix.

This is a rather large refactoring. The "lookup traversal" code was
splitted out from the setup_joins. There is now names_to_path() method
which does the lookup traveling, the actual work of setup_joins() is
calling names_to_path() and then adding the joins found into the query.

As a side effect it was possible to remove the "process_extra"
functionality used by genric relations. This never worked for left
joins. Now the extra restriction is appended directly to the join
condition instead of the where clause.

To generate the extra condition we need to have the join field
available in the compiler. This has the side-effect that we need more
ugly code in Query.__getstate__ and __setstate__ as Field objects
aren't pickleable.

The join trimming code got a big change - now we trim all direct joins
and never trim reverse joins. This also fixes the problem in #10790
which was join trimming in null filter cases.

The trim argument was used by split_exclude() only to trim the last
join from the given lookup. It is cleaner to just trim the last part
from the lookup in split_exclude() directly so that there is no need
to burden add_filter() with the logic needed for only split_exclude().

The code was dead-code since 2006.

Refs #19368, and the fix introduced in 27f8129d.

Thanks to un33k for the report.

This makes it easier to make a ModelBackend-compliant (with regards to
permissions) User model.

Thanks to cdestigter for the report about the relationship between
ModelBackend and permissions, and to the many users on django-dev that
contributed to the discussion about mixins.

Thanks to gcc for the report.

This documents the behavior introduced by cc337a74, which is BACKWARDS
INCOMPATIBLE for any attempt to invoke a method on a manager using the
abstract class as the calling class (e.g., AbstractBase.objects.do_something())

Thanks to mhsparks for the report.

Thanks Hourann Bosci for the patch.