From: | Peter Eisentraut <peter_e(at)gmx(dot)net> |
---|---|
To: | Peter Geoghegan <pg(at)heroku(dot)com> |
Cc: | Matthew Kelly <mkelly(at)tripadvisor(dot)com>, "pgsql-hackers(at)postgresql(dot)org" <pgsql-hackers(at)postgresql(dot)org>, Matthew Spilich <mspilich(at)tripadvisor(dot)com> |
Subject: | Re: Collations and Replication; Next Steps |
Date: | 2014-09-17 18:20:04 |
Message-ID: | 5419D0D4.3000401@gmx.net |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers |
On 9/17/14 2:07 PM, Peter Geoghegan wrote:
> On Wed, Sep 17, 2014 at 11:05 AM, Peter Eisentraut <peter_e(at)gmx(dot)net> wrote:
>>> We could at least use the GNU facility for versioning collations where
>>> available, LC_IDENTIFICATION [1].
>>
>> It looks like the revisions or dates reported by LC_IDENTIFICATION
>> aren't ever updated for most locales.
>
>
> That's not surprising. There is zero controversy about how to
> correctly sort English text, for example. For other languages, that
> might be much less true.
>
a) There is plenty of controversy about how to sort English text on
Stack Overflow. ;-)
b) Even in an English locale you have to maintain a sort order for all
Unicode characters, and that changes more than zero times. But it's
quite clear from looking at the glibc git logs that no one is
maintaining these version numbers.
From | Date | Subject | |
---|---|---|---|
Next Message | Peter Geoghegan | 2014-09-17 18:22:33 | Re: Collations and Replication; Next Steps |
Previous Message | Peter Eisentraut | 2014-09-17 18:15:37 | Re: Collations and Replication; Next Steps |