Encoding nightmare! Pls help!

I've had a discussion on the general list about the implications for
storing accented characters within a postgres (7.4.1) db.

As a result, I have created a database with no locale = C locale (using
initdb {other parms} --no-locale)

Here's the database (test) with UNICODE encoding

List of databases
Name | Owner | Encoding
--------------+----------+----------
test | postgres | UNICODE
template0 | postgres | UNICODE
template1 | postgres | UNICODE

The problem I'm having is that I CANNOT write accented characters into the
database or get them out correctly using my java code and the
pg74.1jdbc3.jar file.

Using psql, I select the data (with client encoding = UNICODE), and I get

tést.jpg

With client encoding = LATIN1, I get

tést.jpg

But in my little java test app, I get:

tést.jpg, tést.jpg, tést.jpg,

I want tést.jpg!!!!!

Here is the offending section of code:

String filename = rset.getString(2);
System.out.print(filename);
System.out.print(", ");

if (filename != null)
{
try
{
filename = new String(rset.getBytes(2), "UTF-8");
}
catch (UnsupportedEncodingException e)
{
System.out.println("Cannot decode string?");
}

System.out.print(filename);
System.out.print(", ");

try
{
filename = new String(rset.getBytes(2), "ISO-8859-1");
}
catch (UnsupportedEncodingException e)
{
System.out.println("Cannot decode string?");
}

System.out.print(filename);
System.out.print(", ");
}

Can anyone explain what I am doing wrong? I have become so confused by all
this, that I don't think I can see the problem straight anymore.

How can I read and write unicode chars into the db. Is there some magick
parameter that needs to be passed when setting up the connection/driver?

Thanks for any/all help!!

John Sidney-Woollett