I used both the auto-generated capi and the specdocs's test forms to send an standard string with japanese text as "text" to the server (Tomcat 5.5.15, OC4J (which just works with the capi)).

Both, the specdocs's test forms and the capi transform the Japanese string into the UTF-8 format specified by W3C at http://www.w3.org/International/O-URL-code.html as specified in org.xins.common.text.URLEncoding.

As for that, the client works fine in my opinion.

However, as the character encoding is not set in the APIServlet of xins 1.4.0, an URL like " http://127.0.0.1:8080/service/?_function=TestFunction&_convention=_xins-std&email=aaa%40aa.aa&name=%E3%81%82"
will be converted from (default) Unicode (or ShiftJIS, not sure) to UTF-8. Thus, the logic gets some strange characters as input from the parameter name. That's because actually the URL is correctely UTF-8 and thus the request's character encoding should be set to UTF-8 before any attempt to get parameters with getParameter.

Thank you as well, Anthony. Although, I think there is no problem with capi or test forms, this setting change could be helpful to enter Japanese characters directly into the address bar.