Re: Add standard collation UNICODE - Mailing list pgsql-hackers

From Laurenz Albe
Subject Re: Add standard collation UNICODE
Date
Msg-id a2f03d437a38c265b25284232098626a8776320e.camel@cybertec.at
Whole thread Raw
In response to Re: Add standard collation UNICODE  (Jeff Davis <pgsql@j-davis.com>)
List pgsql-hackers
On Thu, 2023-03-23 at 13:16 -0700, Jeff Davis wrote:
> Another thought: for ICU, do we want the default collation to be
> UNICODE (root collation)? What we have now gets the default from the
> environment, which is consistent with the libc provider.
>
> But now that we have the UNICODE collation, it makes me wonder if we
> should just default to that. The server's environment doesn't
> necessarily say much about the locale of the data stored in it or the
> locale of the applications accessing it.
>
> I don't have a strong opinion here, but I thought I'd raise the issue.
>
> By my count, >50% of locales are actually just the root locale. I'm not
> sure if that should matter or not -- we don't want to weigh some
> locales over others -- but I found it interesting.

I second that.  Most people don't pay attention to that when creating a
cluster, so having a locale-agnostic collation is often better than
inheriting whatever default happened to be set in your shell.
For example, the Debian/Ubuntu binary packages create a cluster when
you install the server package, and most people just go on using that.

Yours,
Laurenz Albe



pgsql-hackers by date:

Previous
From: Peter Eisentraut
Date:
Subject: Re: ICU locale validation / canonicalization
Next
From: Tatsuo Ishii
Date:
Subject: Re: [EXTERNAL] Support load balancing in libpq