> > Now (assuming the above picture is correct), if we step back back
> > and look at the words, in essence what we have is the interesting
> > situation where the printer can actually *map* a warning condition
> > to a critical condition.
>
> Actually, we have a situation where the printer has been told "stop
> printing when the toner goes low because the user doesn't want to take a
> chance on running out of toner." If the printer truly mapped the toner low
> to a critical alert, it would never print with a low toner condition.
Reading that last sentence, I'm not sure whether you're agreeing
or disagreeing with me. If the printer "would never print with a
low toner condition", then this definitely represents a critical
(ie, show-stopping) situation, right? (Are we agreeing here? Or
am I missing a point?)
> > As such, for this scenario, a "toner low" condition is not at all
> > a non-critical (ie, warning) condition, but rather a very real
> > critical condition. There is nothing in the alert code tagged for
> > "toner low" that implies it is implicitly a non-critical alert, right?
>
> Toner low is implicitly a non-critical alert because technically, the
> printer can still print with no degradation.
Ah, here is where we definitely disagree. (And I believe this might
be just what Bill Wagner was complaining about regarding the attempt
to "standardize" critical vs. non-critical conditions.)
See if you can agree with this series of premises and the resulting
conclusion:
1. Any condition for which the printer ceases operation is described
as being a "critical" alert. (Conversely, a condition can NOT be
described as a "critical" alert if the condition does NOT result
in the printer stopping operations.)
2. Alert codes do NOT have any associated semantics (implied or
otherwise) regarding the severity level of the condition associated
with the alert code. In other words, there is absolutely no
binding whatsoever between an "alert code" and the "severity level"
for any given condition.
Therefore, if the printer ceases operation (ie, stops printing) when
a "toner low" condition is encountered, then the alert added to the
Alert Table must be tagged with a severity level of "critical".
And, following Bill Wagner's relevant comments, whether the printer
chooses to take itself offline or not is completely up to the printer
manufacturer; however, if it does, then it must also add a separate
alert to the table with an alert code of "Offline" and a severity level
of "critical".
Do you agree with this?
By the way, I'm not sure this statement is completely correct:
> Toner low is implicitly a non-critical alert because technically, the
> printer can still print with no degradation.
Until the printer runs *out* of toner, printing *can* continue; however,
at some point those sheets are going to start getting a bit dim, don't
you think? ;-)
> > If this is indeed true, then we don't have to worry about muddying up
> > the Top 25 Conditions table with a new situation in which a non-critical
> > alert is treated in some way as a critical alert.
>
> We aren't muddying up the table. We are just explaining a case where a
> non-critical condition can cause a secondary critical condition. When the
> critical condition is fixed (the printer is put online), the original
> non-critical condition (toner low) remains.
Sure, I can follow your drift here. However, after reading Gail Songer's
just-posted message on this thread, I think I must disagree with some of
the directions we set during the telecon regarding the usage of the
"Offline" alert.
Please read my response to Gail's message (coming out Real Soon Now).
> > Basic premise for everyone to agree upon (and maybe a useful bit of
> > clarifying text for the new Printer MIB draft):
> >
> > The printer (and only the printer) decides which alert codes are
> > critical versus non-critical.
>
> I can agree to the above.
Great. Does anyone else out there in PMP-land disagree with this premise?
> > Does everyone agree with this? If so, then (following Bob's text) when
> > the user "continues" the printer (ie, acknowledges the warning condition),
> > then shouldn't the printer remove *both* the Offline and critical alert
> > (describing the toner low condition), and add a new non-critical alert
> > to say that the condition still remains (but is no longer critical)?
>
> If the printer does this, then it appears as if the toner low condition was
> fixed and then occurred again, when in reality the toner situation hasn't
> changed. Also, this causes the management application to handle yet another
> alert (three instead of two).
Yes, I'll admit I fully see your point. And Gail's message seems quite
related to this line of thinking, too.
I'll address this in the reply to Gail's message.
...jay
PS: We're really coming down the wire on nailing down the Printer MIB,
and this issue is definitely part of that effort. Other PMP folks
are strongly encouraged to voice their opinions to the list so that
we can quickly come to consensus. It makes no difference to me if
you agree or disagree with my statements, but rather that you state
what you think is correct...and quickly, please.
PPS: Now, about not caring whether you disagree with me... ;-)