Hi Juergen,
I agree with you that 'prtMarkerLifeCount' SHOULD be persistent.
And that's what RFC 1759 (Printer MIB v1) said in the same words
used in the latest Printer MIB v2 draft, but in section 2.3, not
in the object definition itself.
Back in 1994, the chance that a low-end printer might have trouble
supporting persistent counters was probably much higher.
I would suggest that you may actually safely assume that the
'prtMarkerLifeCount' _is_ persistent, because I've never heard
of a modern printer where it is not persistent.
Should we improve the DESCRIPTION clause of 'prtMarkerLifeCount'
to explicitly add something like:
"Note: This object SHOULD be implemented as a persistent
object with a reliable value throughout the lifetime of
the printer."
Cheers,
- Ira McDonald, co-editor of Printer MIB v2
-----Original Message-----
From: Juergen Schoenwaelder [mailto:schoenw@ibr.cs.tu-bs.de]
Sent: Wednesday, February 20, 2002 5:03 AM
To: imcdonald@sharplabs.com
Cc: pmp@pwg.org; Ron.Bergman@Hitachi-hkis.com; bwijnen@lucent.com
Subject: Re: question about the printer mib
>>>>> McDonald, Ira writes:
Ira> So there is no absolute guarantee that 'prtMarkerLifeCount' is
Ira> always accurate for the life of the printer.
So what is the practical value of this zero-based counter
prtMarkerLifeCount if management applications can not rely on it?
I would feel much better if the RFC says "if you implement
prtMarkerLifeCount, then it has to be persistent" rather than
allowing non-persistent prtMarkerLifeCount implementations that
basically report some random numbers (wrt. to the semantics of
prtMarkerLifeCount) and potentially confuse applications.
/js
-- Juergen Schoenwaelder University of Osnabrueck <schoenw@inf.uos.de> Dept. of Mathematics and Computer Science Phone: +49 541 969 2483 Albrechtstr. 28, 49069 Osnabrueck, Germany Fax: +49 541 969 2770 <http://www.informatik.uni-osnabrueck.de/>
This archive was generated by hypermail 2b29 : Wed Feb 20 2002 - 16:03:32 EST