Re: Change initdb default to the builtin collation provider

Robert Haas wrote:

> To be honest, I'd probably be ready to support making the default
> encoding UTF8 regardless of the environment, and you have to use -E
> if you want anything else. I think there are still people using
> other encodings, but I believe it to be a small minority at this
> point.

It would be interesting to have the point of view of Asian users about
this. Recently, the suggestion to retire GB18030 in favor of UTF-8 was
met with the objection that GB18030 was likely preferred by users from
China [1]. Another example against UTF-8 that I found notable, is
Tatsuo Ishii mentioning that Japanese users tend use --no-locale
rather than UTF-8 locales [2].

Also, it's not obvious how initdb could choose an UTF-8 locale
regardless of the environment.
For instance, let's say it finds LC_ALL="fr_FR(dot)iso885915(at)euro", what
would it do? Maybe look at the UTF-8 locales on the system. Here's a
subset of what it would find on my system:

C.utf8
en_AG
en_AG.utf8
en_AU.utf8
en_BW.utf8
en_CA.utf8
en_DK.utf8
en_GB.utf8
en_HK.utf8
en_IE.utf8
...
tr_TR.utf8

From that kind of list, which locale should it pick and why?

Personally I think that ignoring the environment's LC_* for the
collations would be fine if we went for builtin/C.UTF-8 by default, as
$subject suggests. But the level of enthusiasm for that from the
community seems much lower than it would need to be for that kind of
change to be acceptable.

[1]
https://www.postgresql.org/message-id/45b4b689-0e78-4d30-a5f9-1a39d01ab2b7%40ww-it.cn
[2]
https://www.postgresql.org/message-id/20230608.104535.2171011311090815110.t-ishii%40sranhm.sra.co.jp

Best regards,
--
Daniel Vérité
https://postgresql.verite.pro/