Another SMIv2 question

["Robert Moore" <remoore@us.ibm.com>]

From RFC 2578, page 25:

>A requirement on "standard" MIB modules is that
>the Counter64 type may be used only if the
>information being modeled would wrap in less
>than one hour if the Counter32 type was used
>instead.

It appears that the recent exchange on the diffserv
list, and the -15 DiffServ MIB that came out of it,
has had the effect of quietly overriding this very
explicit statement in SMIv2.  First, am I reading
this result correctly?  It certainly looks like the
DiffServ MIB has plenty of Counter64's that will not
exceed 32 bits within one hour.

Second, while I have no objections at all to this
new way of using Counter64's, I do worry about these
silent, de facto amendments to the SMI.  We can say
we'll "get it all right" in sming, but meanwhile,
how is someone supposed to know what the rules are
for defining a semantically valid MIB module?

Regards,
Bob

Bob Moore
Advanced Design and Technology
Application Integration Middleware Division
IBM Software Group
+1-919-254-4436
remoore@us.ibm.com