Comment 3 for bug 1284660

Vanh, thank you for the verbal briefing on this bug. My general comment would be that when decoding section 1, the value specified in octets 1-3 is the only "correct" value for the length of section 1. In a correctly encoded BUFR message, the total number of octets in section 1 must match the length given. A message that does not meet this condition is malformed and should trigger an error condition. And finally, there is no requirement for section 1 to be of even length.