Forum OpenACS Development: Response to PostgreSQL upper() / lower() functions and LATIN2 charset

No. When creating database, you should use LATIN1 encoding - LATIN2 is
Central European (for most Slavic countries), while LATIN1 is
appropriate for Germany, France etc.

so it should be:
CREATE DATABASE test WITH ENCODING='LATIN1';

in psql, you should use LATIN1 encoding. You should use the UNICODE
client encoding inside AOLServer - and that should do the trick (this
is done by setting PGCLIENTENCODING env. variable, though I asked Don
Baccus to add this function to the PG driver).

All this may not be neccessary, if using 2.2 glibc and 2.4 kernel
(though I haven't tested such configuration yet)