RE: RUNT Packets
> By the way,
> the traditional criteria in the IETF for including an item in the MIB is
> way below our 75% threshold and has more to do with whether anybody has
> implemented it than with it is the industry consensus to include it.
>
[Dan] Geoff,
Tripathi mentioned in a previous mail that several vendors actually
provide the RUNT counters in their MAC implementation, I am wondering if we
are not evading the problem and opening the door to non-interoperable
implementations by not defining the counter someplace in the 802.3. (I
understand the PAR scope issue for ae!)
BTW, you provided a rather 'free' interpretation of the IETF
criteria.
There are three degrees of advancement on the IETF standards track
(see RFC 2026 for more details):
* the entry level is Proposed Standard - the specification is supposed
to be stable, well-understood and having solved the known design choices. It
went through a broad community review. The process is different in form, but
not in substance (IMO) from the one followed in the IEEE editing and
balloting.
* The second level is Draft Standard - at this stage a successful and
documented operational experience is required, and at least two
implementations from different sources must have been proved as
interoperating. At this stage, for a MIB implementation, the MIB objects
that meet the criteria must have been implemented in the field, and their
definition proved to be accurate enough to allow different implementations.
Objects that de not meet these criteria are deprecated, but they cannot be
replaced, and their semantics may not be modified.
* A Full Internet Standard is a specification for which significant
implementation and successful operational experience. Only widely deployed
technologies make it to this stage. Interestingly enough, the RMON MIB is
one of the very few MIB documents that reached this status.
Including an item in a MIB document targeting the IETF standards
track would be a decision for the relevant Working Group to take in the
first phase of the design of the MIB. Criteria for decision would include
the need for such an object and the price for its implementation.
Implementation in commercially available MACs would raise the chances for
inclusion of the object in the IETF MIB, but a standard reference in an IEEE
document is always preferable.
Regards,
Dan