#35194: Postgres 16.2 with _iexact leads to IndeterminateCollation
-------------------------------------+-------------------------------------
Reporter: Aldalen | Owner: nobody
Type: Bug | Status: new
Component: Database layer | Version: 5.0
(models, ORM) |
Severity: Release blocker | Resolution:
Keywords: | Triage Stage: Accepted
Has patch: 1 | Needs documentation: 0
Needs tests: 0 | Patch needs improvement: 0
Easy pickings: 0 | UI/UX: 0
-------------------------------------+-------------------------------------
Comment (by Simon Charette):
It is effectively a solution but I'm not convinced this will do more good
than harm.
#3575 has been merged 16 years ago this means in between now and then
thousands of projects were created and added a functional index on
`UPPER("col")` to make `i(exact|contains|startwith)` use an index and the
moment they upgrade to a minor version of 5.0 their database will start
running slow queries as their indices will be unsuables.
On the other hand we have a bug in a newly introduced feature for a very
particular use case that might be affecting only a few users (must use
generated field, must be on a latest version of Postgres, must use
`i(exact|contains|startwith)`.
I appreciate the intent to solve this issue but I think we need to dig
deeper to truly understand ''why'' this is happening before jumping to
conclusions here as there are no true urgency to get things right here;
the ''release blocker'' assignment is self-imposed and nothing prevents us
from deferring a solution to this problem to a future 5.0 release if we
can't understand why this is happening before the May release as for all
we know if might be a bug in Postgres itself.
I tried reaching out on libera.chat#postgres IRC to get an answer but no
one could answer me their (first time this happens) so I was planing to
reach out to their mailing list this week but I might run out of time so
if someone feels comfortable doing so please do.
To summarize I think we should understand why this is happening before
taking any potential harmful action here. For all we know many other
functions and lookups could be affected and this is just the tip of the
iceberg.
--
Ticket URL: <https://code.djangoproject.com/ticket/35194#comment:10>
Django <https://code.djangoproject.com/>
The Web framework for perfectionists with deadlines.
--
You received this message because you are subscribed to the Google Groups
"Django updates" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To view this discussion on the web visit
https://groups.google.com/d/msgid/django-updates/0107018ef121a8ca-7c7d6268-be24-4287-8d76-7eb808070607-000000%40eu-central-1.amazonses.com.
