Everything in OpenID is supposed to use UTF-8 character encoding, right? Has anyone else run into any encoding issues when validating signatures? (For example, when creating the key-value list of signed fields, or when doing HMAC-SHA1 functions.)