Isn't this behavior actually a bug that should be fixed rather than
documented?
These JSON path functions are specified by the SQL standard, so they
shouldn't depend on PostgreSQL-specific settings. At least in new
functionality we should avoid that, no?
On 10.09.24 21:43, David E. Wheeler wrote:
On Sep 10, 2024, at 14:51, Tom Lane <t...@sss.pgh.pa.us> wrote:
Pushed with a little additional polishing.
Thank you! Do you think it’d be worthwhile to back port to 17?
I thought the best way to address jian's complaint about DateStyle not
being clearly locked down was to change horology.sql to verify the
prevailing setting, as it has long done for TimeZone. That's the
lead test script for related stuff, so it makes the most sense to
do it there. Having done that, I don't feel a need to duplicate
that elsewhere.
Yeah, that will help, but I still bet next time I go to figure out what it is
I’ll stick that line in some test to make it fail with clear output for what
it’s set to 😂.
