>This MIB imports the BITS construct, which should not be done. The BITS
>definition has its problems, but taken reasonably literally its definition
>is available implicitly by IMPORTing OBJECT-TYPE from SNMPv2-SMI.
I disagree. Why should BITS not be imported? It is a textual
convention defined in RFC1902. As far as I can tell, the BITS clause
adds semantic clarity to the notion of assigning meaning to bits in an
octet string.
Bobby