Re: pg_dump and money type - Mailing list pgsql-hackers

From Alvaro Herrera
Subject Re: pg_dump and money type
Date
Msg-id 20070912013540.GD16107@alvh.no-ip.org
Whole thread Raw
In response to Re: pg_dump and money type  (Tom Lane <tgl@sss.pgh.pa.us>)
List pgsql-hackers
Tom Lane wrote:
> "Heikki Linnakangas" <heikki@enterprisedb.com> writes:
> > Tom Lane wrote:
> >> This risk seems rather overstated, as it's unlikely that someone using
> >> money would choose to reload their data into a DB with a fundamentally
> >> incompatible locale setting.  
> 
> > It doesn't sound unlikely at all to me. For example, people often use
> > C-locale for performance reasons, or because of ignorance of locale
> > issues. One scenario that seems particularly likely is to initialize and
> > load a database with en_US or C locale, and run like that for a few
> > weeks. After that, you notice that something's wrong, strings are sorted
> > in a funny way, etc. You realize that you're using the wrong locale, so
> > you take a backup with pg_dump, re-initdb with correct locale, and restore.
> 
> If you're using type money, you will certainly have noticed whether it
> spells the currency sign the way you like.  I can believe that someone
> might go for a while with C where they should have used en_US, or vice
> versa, but not that they'd have failed to notice the difference between
> $ and DM, say.

Well, that assumes that countries use different signs.  We for example
use $, as does Argentina (but they add two decimal places where we have
none), Mexico and US.  And while both Arg. and Mexico use $ and 2
decimal places, the former uses "," for decimals while the latter
follows the US lead and uses ".".

-- 
Alvaro Herrera                                http://www.CommandPrompt.com/
PostgreSQL Replication, Consulting, Custom Development, 24x7 support


pgsql-hackers by date:

Previous
From: Tom Lane
Date:
Subject: Re: pg_dump and money type
Next
From: ITAGAKI Takahiro
Date:
Subject: CVS HEAD is broken by flex