RE: draft-ops-endpoint-mib-04.txt

[Bill Fenner <fenner@research.att.com>]

>On your second point though, if there are multiple addresses in an INDEX
>clause then you cannot rely on the number of OIDs remaining.

IMPLIED can only be present for the last object in the INDEX clause,
therefore if there are multiple addresses all but the last must have
the length included.

I guess I'm being paranoid about OID length but focusing on the wrong
place - IMPLIED can at best save a single ID since it can only be used on
the last object.  Even though tables with a single InetAddress in their
INDEX can get away with IMPLIED, that's not likely to be the tables where
OID length is important.  So, I can be swayed by the consistency argument.

  Bill