Data entered in forms is sent back as URIs or as a URL-encoded body
Problem: Reliable character encoding identification
Various provisions, none of them fully established:
-
Use encoding identical to document received
-
Use hidden field to track transcodings
-
Use body part or file upload (multipart/form-data, RFC 1867)
-
Use accept-charset attribute on
<form>
-
Internationalize URIs with UTF-8
