Re: Question on: draft-ietf-ipv6-rfc2013-update-04.txt

["David T. Perkins" <dperkins@dsperkins.com>]

HI,

Bill - showing exactly the status in a definition in a MIB
module does conclusively provide the authoritative answer
to "what is the status of a definition". Also, even more
importantly, when an item's status is changed to "deprecated"
or "obsolete", the DESCRIPTION clause MUST BE updated
to specify why this was done. And when an definition is
replaced with another, then the DESCRIPTION text MUST
identify the replacement.

Also note that in many cases definitions SHOULD be changed
to obsolete and not depreciated. The meaning of deprecated
in SNMP MIB modules is different than the meaning used
in other technical specifications!!!!!!
 
Many people in the SNMP community have bought into the
incorrect mantra of "first you deprecate and then you obsolete".
This is wrong. You change status for a particular reason
(which you say in the DESCRIPTION clause). This seems
like "extra work" for the document editors, but it's
not extra - it's required - to help out all of the
readers and users of the document.

On Tue, 26 Oct 2004, Bill Fenner wrote:
> Our original thinking (I don't want to think about how long ago that
> actually was) was that including the objects as deprecated would just
> bulk up the document with no obvious benefit.  I don't really know how
> NMSes work, since I always roll my own, but my impression is that the
> deprecated/obsoleted stuff is more for people inspecting MIBs deciding
> how to monitor something than for machines.
> 
>   Bill
> 
Regards,
/david t. perkins