Thread: server , client encoding issue

surabhi.ahuja wrote:
> when i connect to it
> it shows: server_encoding : LATIN9
> and client_encoding: LATIN9
>
> from where did it get these values

 From your environment - the machine is presumably defaulting to some
locale that uses LATIN9 for it's character-set.

> i want them to be UNICODE.
>
> please help
> ( i know there is a workaround to this problem -
> /usr/bin/createdb -E UNICODE temp )
>
> but this is happening at a particular machine . On other machines it is UNICODE.

It's not really a workaround - I always specify encoding for the
databases I create (precisely to avoid this sort of problem).

--
   Richard Huxton
   Archonet Ltd

surabhi.ahuja wrote:
> i checked the locale it is giving:
>
>  LANG=en_US.iso885915
> LC_CTYPE="en_US.iso885915"

If you Google for "ISO-8859-15 Latin9" the top two results seem to give
details. Oh - there are two naming systems for character sets, just to
make things even more complicated.

Now, traditionally you'd have used Latin1 (ISO-8859-1), but the
introduction of the Euro meant they needed to introduce a new character.
They took the opportunity to make some other changes too and called the
results Latin9 (ISO-8859-15).

OK - now the original problem was with a database not having a UNICODE
encoding. It does look like this is because the environment on this
machine is Latin9 rather than UTF-8. It's easy to have this problem, and
I always recommend setting the encoding explicitly when creating a
database cluster (initdb --encoding=UTF8). If you installed from a
package, it might have chosen a default for you though.

HTH
--
   Richard Huxton
   Archonet Ltd