The MPLS-OPS Archive

[Mathew Richardson <mrr@nexthop.com>]

> Roger Clark Williams <rogerw@nordlink.com> [Fri, Jul 12, 2002 at 07:27:38PM -0400]:
>Sameer, I had a Worldcom guy in a class once (will I ever again?) and we 
>had a nice back-and-forth on the subject of standards: If a company only 
>supports a subset of an RFC or an IEEE standard, is that company complying 
>with the standard? His argument was no, mine was yes. There is no right or 
>wrong answer here.

<snip>

First, let me say that as long as an implementation satisfies the
requirements behind the decision to use a standard, then the question
of whether or not an implementation actually complies with the
standards, or not, is irrelevant.  For example, if a standard is
adopted exclusively to promote interoperability, and an implementation
is 100% interoperable with all past, present, and future compliant
implementations, then it doesn't matter whether the implementation
is actually compliant with the standard or not.  The goals specified
when adopting the standard have been achieved.

Having said that, there _is_ a right or wrong answer when it comes
to determining whether an implementation is in compliance with a
standard: if an implementation conforms with all of the "required"
elements of a standard (the MUSTs and MUST NOTs, for IETF standards),
then it is in compliance; otherwise, it is not.  So, if a company
supports only a subset of a standard, then whether it is in compliance
with that standard, or not, depends on which subset is supported.

mrr

-------
The MPLS-OPS Mailing List
Subscribe/Unsubscribe:  http://www.mplsrc.com/mplsops.shtml
Archive: http://www.mplsrc.com/mpls-ops_archive.shtml