Looks like Tom put this all in.
---------------------------------------------------------------------------
Magnus Hagander wrote:
> The following patch fixes locale support under win32.
>
> * Saves and reloads LC_COLLATE and LC_CTYPE when a new backend is
> execed. Also preserved in pgstat even though it's supposedly not used
> there at the moment, to be on the safe side for the future. With this
> patch, passes regression tests with regional settings set to != english.
>
> * This is not necessary (see previous discussion) to do for the other
> LC_xxx types. These are all set in GUC, and are handled by the routines
> to save/restore all nondefault GUC varilables.
>
> * setlocale(LC_MESSAGES,"") returns NULL on win32. And it's always
> called as part of GUC setup. Patch changes code to ignore the result if
> we set it to "". (Can someone confirm that this is not totally unsafe?
> From what I can tell, LC_MESSAGS is only set to ""
>
>
> Oh, and how do I change a locale:d system into english? Mine picks up
> swedish from the environment...
>
>
> After this, the only remaining regression error on win32 is the join one
> (the "unimportant ordering" one)
>
> //Magnus
>
Content-Description: locale_win32.patch
[ Attachment, skipping... ]
>
> ---------------------------(end of broadcast)---------------------------
> TIP 6: Have you searched our list archives?
>
> http://archives.postgresql.org
--
Bruce Momjian | http://candle.pha.pa.us
pgman@candle.pha.pa.us | (610) 359-1001
+ If your life is a hard drive, | 13 Roberts Road
+ Christ can be your backup. | Newtown Square, Pennsylvania 19073