Hi Jay and PWG folks,
Xerox has an active (currently private) MIB which is a registry
of product codes as 'autonomous OIDs' which conform to the
'ProductID' textual convention in RFC 1514, for use in
'hrDeviceID' (and possibly also 'sysObjectID' in MIB-II
for low-end systems - we know about the fact that rigorously
the MIB-II attribute should only identify the version of
the SNMP Agent in the managed system and NOT the whole
system product).
As Xerox products begin to ship with active support for
meaningful 'hrDeviceID' (fairly soon), Xerox will make
publicly available this 'Product ID MIB'. If Underscore
would like to maintain a 'collection' of these MIB-lets
for Printer MIB compliant vendors, that seems like a
fine idea.
On a related thread, per RFC 1573 and common implementations
the 'ifTable' in MIB-II DOES show so-called non-network
interfaces (serial and parallel, point-to-points).
Xerox developers ARE showing all interfaces. The spot
that is cloudly is that RFC 1514 separated out three
'hrDeviceType' values 'hrDevice[Network|Serial|Parallel]'
and said only 'hrDeviceNetwork' ones fill in the
'hrNetworkTable' (pointer to 'ifTable' via 'ifIndex'
value).
I've seen 'parallel' interfaces (in 'ifTable') shown
via the 'prtChannelTable' in shipping printers - we
certainly intend to do so at Xerox. PLEASE don't
get stuck on some loose definition of 'interface'.
A source of job datastreams is a source of job
datastreams is a...
Cheers,
- Ira McDonald
(outside consultant at Xerox for network management)
High North Inc
PO Box 221
Grand Marais, NY 49839
PS - Jay - Nancy and I leave for ten weeks in New Zealand
on 7 Nov 96 - talk to you when we return in late Jan 97!