On 6/5/23 18:07, Daniel Verite wrote:
While trying pg16beta1 libc collations on Windows, I noticed that UTF-8 text sorts sometimes differently across invocations with the same locales, which is wrong since these collations are deterministic.
<snip>
Also, it does not occur at all if parallel scan is disabled.
This is a wild shot in the dark, but I wonder if somehow the locale is being initialized (i.e. setlocale) differently in the parallel workers than the backend due to some Windows specific behavior differences?
-- Joe Conway PostgreSQL Contributors Team RDS Open Source Databases Amazon Web Services: https://aws.amazon.com
