RE: Updating the MIB boilerplate

["Andrew Smith" <ah_smith@acm.org>]

Specifically, I'd hope that we're still trying to maintain the (almost
complete) separation of the protocol (e.g. SNMPv3) from the description
language (i.e. SMIv2, since it's the boilerplate for an SMIv2 MIB that
we're writing here). Perhaps the boilerplate is the place where we ought
to hint at this imperfect separation and call out any areas where
protocols (limit ourselves to discussion of SNMP, since that is the only
IETF-sanctioned protocol that uses an SMI of this version) are known to
impact the use of the things in the MIB (e.g. the lengths of some string
variables - other areas too?). But please cut out any sort of historical
stuff about the protocol, as opposed to the language.

My 2c,

Andrew Smith


-----Original Message-----
From: owner-mibs@ops.ietf.org [mailto:owner-mibs@ops.ietf.org] On Behalf
Of Juergen Schoenwaelder
Sent: Tuesday, October 29, 2002 9:50 AM
To: heard@pobox.com
Cc: mibs@ops.ietf.org
Subject: Re: Updating the MIB boilerplate



>>>>> C M Heard writes:

Mike> On Tue, 29 Oct 2002, Juergen Schoenwaelder wrote:
>> Since SNMPv3 is full standard now, I hope we can actually shorten
>> the boilerplate significantly.

Mike> That does not strike me as wise. 

[...]

This goes back to an old discussion whether we really need to explain
the various versions of SNMP and the SMI in great detail in each MIB
document. Sure, repetition is the key to learning... but perhaps at
some point in time, we can simply stop with this education by pointing
to RFC2570bis which has all the details.

In order to understand a MIB module, you basically need to know the
SMIv2 (normative) but not all the history about the SNMP versions.

Anyway, lets see what other feel about this old topic at this point in
time...

/js

-- 
Juergen Schoenwaelder
<http://www.informatik.uni-osnabrueck.de/schoenw/>