Classification:
One the phone: (pardon misspelled names)
Ira McDonald
Jay Martin
Bob Pentecost
Chuck Adams
Ron Bergman
Matt King
Lloyd Young
David Kellerman
Harry Lewis
Hope I didn't miss anyone.
Meeting was guided by Lloyd's 4/48 agenda
First item covered was "Old Item (1)" the Top 25 alerts. Reviewed Chuck's table
and Bob's review comments.
- We decided that we should call this task the Top 25 (21?)
conditions, not alerts. Not every condition we're dealing with
is an alert.
- Big discussion about how to define Offline. As usual, we did not
reach consensus on how to DEFINE off-line, but we did agree that
off-line should be treated as a separate condition. We decided
that an off-line alert table entry should reference GroupIndex 5
(the General Group) with subUnitIndex (-1).
- We decided that markerSupplyLow (or missing) conditions alert table
entries should reference the MarkerSuppliesGroup (11) and not the
marker group. Marker supplies, however is indexed to a marker.
Subunit status should reflect the correct marker for the supply
that went low or is missing.
- Bob Penetcost proposed an excellent method for handling the toner
low with continuation (button) scenario. Bob will write up and
submit the proposal.
- We agreed to remove references to old prtAlertCodes when they have
been superseded by the new generic codes.
- Chuck Adams recorded finer details than what are represented in
these minutes and will modify and post the updated table.
Next, we addressed Matt King's clarifications which are in his 4/21
e-mail.
- Matt's item 1a (Clarification of redundant unary events) will be
removed. It will be up to the implementation when to remove
redundant unarys (when the latest unary is added or with general
table aging and management). This satisfies an objection Bill
Wagner had made to Matt's proposal to always manage out redundant
unarys.
- Matt's item 1c (managing persistent binary alerts back into a
crowded table) was accepted with the emphasis still on making
your table large enough that critical events should not get
managed out in the first place. Also, there is a clarification
that, by definition, critical binaries MUST be managed back in
when other conditions are cleared and warning binaries can be
managed back into a crowded table (i.e. toner low).
- Matt's item 3 (clarification of prtChannelCurrentJobCntlLangIndex
and prtChannelDefaultPageDescLangIndex) index = 0 if none was
accepted. It was further emphasized that prtAlertGroupIndex and
prtAlertSubUnitIndex are -1 and will remain that way under similar
conditions.
- Matt's item 4 addressed what constitutes a config change. Both Matt
and Harry had submitted "lists". Jay had written a "philosophy".
We concluded a list could always be wrong but a philosophy could
never be misinterpreted ;-) so it is better to stick with Jay's
paragraph which basically says you either have an ATTRIBUTE or
NOT (NOT = Level, Gauge, Status or Counter). If an ATTRIBUTE
changes, that's a config change. Obtuse example is change in the
description of the Output bin. Even THIS should bump the config
change counter.
- Harry gets to clarify the prtInterpreterFeed and XFeed
Addressability (I wonder why?).
- Ron's draft regarding IANA registry of Interfaces was accepted
- Chuck's PDF enum definition was accepted.
- We seemed to buz through the rest due to short time and we seem
to have left openings for some further discussion even though
we only have until Friday to get final comments to Randy.
Harry Lewis - IBM Printing Systems