Hi Gunnar,
"Gunnar Groetschel" <ggroetschel@sokoma.de> writes:
> If i use the odbc driver i see only sh** (Möller should be Müller).
That looks like UTF-8.
> psql -l shows me, that the database is in SQL_ASCII (what should be
> ok - after 2 hours reading manual pages).
I haven't read the manual for this, but strictly speaking "ASCII"
means, that you don't have umlauts at all, or rather that their
encoding is undefined. Which is not good for a stable database,
because with any update of any tool, that tool can change its internal
default interpretation and potentially start corrupting data. You
really do want a database encoding that supports your data explicitly.
Some ODBC-based tools let you configure how to encode data going in
and out of the ODBC API. Also in the past I have had success with
issuing
SET CLIENT_ENCODING TO 'UNICODE'
on the connection once to get the data in UTF-8. You may want to see,
if a similar statement can be used to change the DB interface encoding
to Latin-1 or Windows-1252.
benny
