A lot of earlier created Oracle database did not configure the character set correctly in Asia, they choose US7ASCII, WE8DEC, WE8ISO8859P1 as the database character set. By default they just keep the NLS_LANG the same as database setting, and then load some Asia language data into the database, no problem under this configuration except when they choose Java language to develop the application. Just cannot get the Asia language displayed correctly.
I create a demo table in a WE8ISO8859P1 database, and test it with AnySQL utility ( a Java JDBC SQL utility, looks like Oracle SQL*Plus) under JRE 1.4 and JRE 1.6 by following two commands.
c:\jdk160\bin\java -Dfile.encoding=ISO-8859-1 -jar jlib\oasql.jar
c:\jdk142\bin\java -Dfile.encoding=ISO-8859-1 -jar jlib\oasql.jar
For JRE 1.6, it can display Asia language correctly.
ASQL> SELECT 'some chinese' as COL1 FROM DUAL;
COL1
----
<display correctly>
1 rows returned.
But it does not work for JRE 1.4, and works only for console base program. Someone asked how to get Oracle SQL Developer work correctly under this situation, but I have no idea of it.