Yes, the original UTF-8 spec was much more forgiving. The most recent spec, RFC 3629, restricted the range of code points and made the decoding of invalid sequ

Yes, the original UTF-8 spec was much more forgiving. The most recent spec, RFC 3629, restricted the range of code points and made the decoding of invalid sequences a MUST NOT requirement: http://tools.ietf.org/html/rfc3629#section-12 http://tools.ietf.org/html/rfc3629#section-12