In my application, for the name field I entered polish characters eg: ŁĄSŻKIĘWIĆZIUŚ and language is polish. I navigated to one tab to the other. The polish characters are getting changed to ?. Please help me to resolve the problem
The only other two significant differernce would be the JVM and the data source helper class. WebSphere will be using an IBM JVM and (usually) wraps all datasource access through a helper class. However, I'm currently looking at at WebSphere/Oracle combination that is showing (what I understand to be) valid Polish characters so I'm assuming these two things are fine. If your application stack uses Unicode through out, and your HTML pages correctly set the charset in a meta tag then this should just work. The only other thing that might cause an issue is if the client you use can't support Polish - but presumably you are accessing your weblogic and webshpere apps from the same client?
Not sure what else to suggest. It sounds like somewhere in your application's string handling you are doing something unsafe, and the difference between the environments is exposing it. Do you use getBytes anywhere? Or the new String(byte bytes) constructor?