Hi,

> Read code in Connection.java in the following places.
>
> private final static String defaultEncoding__ =
> CharacterEncodings._8859_1;
> Connection()....................
> addRequiredPropertiesAndSetConverters()........................

This is encoding used to convert strings to byte[] and back. Default
charset for connection (DPB value) is still NONE (see
ConnectionProperties.java). Maybe Borland is correct that we need to
use ISO8859_1 when converting strings to byte[] when lc_ctype
is "NONE". I will check this.

> I have run the internal tests for the following character sets and
> there are not failures at all.
>
> 1.- Server WIN1252, Client not set
> 2.- Server WIN1252, Client ISO8859_1
> 3.- Server WIN1252, Client WIN1252
> 4.- Server NONE, Client WIN1252
> 5.- Server NONE, Client not set

Did your database contain language specific characters? Have you
tried TestFBEncoding.java test case?

Regards,
Roman