| Age | Commit message (Collapse) | Author |
|
|
|
|
|
|
|
|
|
Deprecates the "--plain" option.
|
|
BaseCommand.execute(**options).
|
|
This includes the ability to collapse inlines by specifying a class named
'collapse'.
|
|
Added unapplied migration count and the list of unmigrated apps.
|
|
|
|
|
|
|
|
|
|
This fixes a regression introduced by a5f6cbce07.
Thanks Gavin Wahl for the report and Tim Graham for the review.
|
|
|
|
Thanks to Trac alias benzid-wael for the report.
|
|
|
|
on Oracle.
|
|
|
|
This fixes a regression introduced by 871440361.
|
|
Thanks to Geoffrey Sechter and the Django NYC group for the report and Markus
for the review.
|
|
|
|
|
|
There is no need to wrap the save() call in transaction.atomic() as
it's already done down the call stack in Model.save_base().
|
|
|
|
Thanks to Trac aliases gerricom for the report, raphaelmerx for the
attempts to reproduce and Sergey Fedoseev and Tim for the review.
Refs #16891
|
|
|
|
|
|
inheritance child models.
|
|
Thanks Ben Kraft for the report, and Simon Charette for the review.
|
|
Thanks to bmispelon and uruz for the initial patch.
|
|
|
|
backwards-compatibility alias.
|
|
not expected.
|
|
|
|
|
|
Thanks hop for the report and patch.
|
|
|
|
AlterField.
Thanks to Emanuele Palazzetti for the help.
|
|
installed.
|
|
|
|
Referencing URLs by dotted path is a stronger convention than
importing the URLs.
|
|
SeparateDatabaseAndState.database_backwards().
|
|
Due to randomness of garbage collection with geometry objects, it's
easier to simply ignore AttributeError/TypeError generally raised when
parts of objects are already garbage-collected.
Thanks Sergey Fedoseev and Tim Graham for reviewing the patch.
|
|
URLValidator now validates the maximum length of a hostname and the
maximum length of all labels inside the hostname.
|
|
method.
|
|
Thanks to Jean Gourds for the report, Tim and Claude for the review.
|
|
|
|
This allows setting a Content-Security-Policy HTTP header
(refs #15727).
Special thanks to blighj, the original author of this patch.
|
|
This also fixes a possible data eviction race condition between
setting and getting a key. Another thread could remove the key
before get_and_set() accesses it again. In this case, now the
default value will be returned instead of None.
|
|
|