We are using mostly automated deployments to quickly provide the latest
changes to our customers. Whenever we update to a newer Django, we need to
clear the whole cache. Automating this requires special handling, as
running the cache invalidation on every deployment seems to be overkill
for unrelated updates - thus some custom logic is required to only force
an invalidation on Django package changes when necessary at the moment.
For this reason, I would like to propose an additional/changed solution
which turns this into hard failures which can be handled by my application
without having to invalidate the cache itself. This way, I am able to
cleanly detect this and handle it appropriately.
Due to being a backwards-incompatible change, this might be changed by a
new setting which defaults to the old behavior.
--
Ticket URL: <https://code.djangoproject.com/ticket/34973>
Django <https://code.djangoproject.com/>
The Web framework for perfectionists with deadlines.
* status: new => closed
* resolution: => wontfix
Comment:
Thanks for the report, however you should be able to make these warnings
errors using `warnings.filterwarnings()`. There is no need to change
anything in Django itself.
--
Ticket URL: <https://code.djangoproject.com/ticket/34973#comment:1>