Hi Ian,
Thanks for your insights.
On Tuesday 21 May 2013 23:22:14 Ian Kelly wrote:
>
> You may be right. One thing I would be concerned about is the
> reliability of only doing implicit date conversions. If you google
> for "ORA-01843" you can find a good number of forum posts that
> recommend against relying on the NLS_TIMESTAMP_FORMAT setting.
I looked at some of those. Most of them warn against it on two grounds:
- The actual format may be different from what you expect
- Formats that use the MON specifier are actually language-dependent
Both of these are mostlly irrelevant for Django -- we set the timestamp format
explicitly, and we don't use MON. A user may change the date/timestamp format
via raw sql, and cause breakage (especially with the new persistent
connections), but -- given the history where we set the format on every
connection -- I think such users mostly deserve what they get. At most, we
should warn against it in the documentation.
> See also Django ticket #20292.
Comment #3 on that ticket suggests that explicit casts do little to improve
things there -- apparently Oracle 10 has a problem applying formats to unicode
strings. I don't have Oracle 10 accessible to test against, and Django's tests
include instances of saving models with DateFields -- so I know the problem
(as reported) is not reproducible against Oracle 11. So as far as I can see,
removing the explicit cast makes things no worse here too.
I just committed the removal in Oracle.
Thanks,
Shai.
--
You received this message because you are subscribed to the Google Groups
"Django developers" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To post to this group, send email to [email protected].
Visit this group at http://groups.google.com/group/django-developers?hl=en.
For more options, visit https://groups.google.com/groups/opt_out.
