Uncommenting the setCharacterEncoding line does nothing to change the result.
I have also tried this: new String(request.getParameter("username").getBytes("iso-8859-1"), "UTF-8");
and setting the page encoding:
<%@ page language="java"contentType="text/html; charset=UTF-8" pageEncoding="UTF-8"%> both with no luck.
Any ideas why the data is converted differently between Tomcat versions, both using the same JRE: 220.127.116.11?