PMP Mail Archive: PMP> Minutes of 4/22 Conference Call

PMP Mail Archive: PMP> Minutes of 4/22 Conference Call

PMP> Minutes of 4/22 Conference Call

Ron Bergman (rbergma@dpc.com)
Wed, 23 Apr 1997 06:58:09 -0700 (PDT)

The minutes of the 4/22 PMP conference call have been posted in

ftp://ftp.pwg.org/pub/pwg/pm/minutes/conf-call-970422.txt .doc .pdf

I have also attached the text version.

Printer MIB Teleconference Minutes
April 22, 1997

Participants:

Lloyd Young
Ron Bergman
Ira McDonald
Dave Kellerman
Bob Pentecost
Gail Songer
Scott Isaacson
Matt King
Randy Turner
Chuck Adams
Harry Lewis
Steve Thorne

CLARIFICATIONS FROM INTEROP TESTING:

The five issues in this group are postponed until next weeks call. The
proposal from Matt (items 1 - 4) was posted late last night and this
will give everyone more time to review. Bill Wagner was not available
for issue number 5.

OLD ITEMS:

1. Top 25 Alert Conditions

Bob Pentecost commented that the "off-line (no binary alerts)" entry
did not show an entry in the alert table. Bob also proposed that
subUnitOffline (22) should be the proper entry in the alert table. It
was generally agreed that this would be a logical result and several
implementations conform today. Bob agreed to present a formal proposal
for this change. Bob also has several more comments on the alert table
which he will also send to the mailing list.

2. RFC1213 references in Printer MIB

Ron Bergman presented the results of his comparison of RFC1573 and
RFC1213. Actually, RFC1573 is an extension of the interfaces group of
RFC1213. Four objects in RFC1213 are deprecated, two of which are
replaced with four new objects.

Ira McDonald pointed out the danger of replacing the reference to the
interfaces group of RFC1213 with RFC1573, since RFC1573 is presently
only a Draft standard. The conclusion was that all we need is the new
interface type enums and these are IANA registered. Ron agreed to draft
new text that references the IANA registry for the complete list of
available interface enums.

3. Should a PDF type be included?

Discussion of this subject was postponed until next week since Chuck
Adams has not posted his proposal.

4. Should the phone:, fax: and email: tags be used in non 8-bit
environments? non-English environments?

Ira discussed his proposal for three Textual Conventions for grouping
all currently defined strings. Randy also discussed his proposal that
the current localization should apply to all read/write strings that are
intended to be human readable and are not explicitly controlled by
prtConsoleLocalization.

After much group discussion, it was agreed that Randy will review
Ira's proposal and send a recommendation to the mailing list. The goal
is to allow the strings to be localized with a minimal impact to the
document.

5. Invalid MAX-ACCESS clause on prtAlertIndex.

Ira has concluded that it is impossible to write a v1/v2 compatible
MIB that has Traps. Ira will post a proposal to the mailing list and
the group can review in next weeks call.

NEW ITEMS:

1. Conformance Major typo (Bob Pentecost 3/21/97):
Included in latest draft.

2. prtGeneralPrinterName (Jay Martin 3/21/97):
Included in latest draft.

3. New IETF Draft Review (Bill Wagner 4/4/97):
Initial proposal modified, new item raised is #4.

4. Novell Channel Type Info (Scott Isaacson 4/16/97):
It was agreed that the purpose of prtChannelInformation was to allow a
client to obtain sufficient information to print. Information
concerning the channel that a management station may find useful is
beyond the scope and intent of this object. Thus Scott proposed that
for PServer the necessary information is File Server Name and Queue Name
for 3.X servers and Tree Name and Queue Name for 4.X servers. The other
information in the recent email was not required for a client to print
and will not be included.

The channel table does not have to be a dynamic table. (i.e. If a
server goes down, the entry for that server can remain. The client must
use some other method to discover the server is down.)

Using Unicode can require a string length of 512 bytes for the PServer
entries. SNMP recommends 255 as a limit to allow for the case of a
minimum packet size. After much discussion it was proposed that we just
accept 255 as the limit and be done with it. Scott agreed to review all
of the information discussed and submit a new proposal.

5. Reference to other MIB's in RFC (Bill Wagner 4/17/97):
It was agreed that this text should be added.

6. Defect report in Printer MIB v2 (Ira McDonald 4/16/97):
See OLD ITEMS #4.

7. prtAlertTime (Bob Pentecost 4/17/97):
What should be the value of prtAlertTime if the agent cannot obtain
any value of time. The agreed response should be zero. After further
discussion it was agreed that Randy should clarify the description of
prtAlertTime to indicate that it shall be equivalent to either sysUpTime
or hrSystemUpTime.

8. Description for chPortHTTP (Ron Bergman 4/17/97):
The requested change will be incorporated.

9. Traps - new info (Chris Wellens 4/17/97):
The discussion will be deferred to the next call. It was agreed that
another test session will not be required. But some additional
independent testing is necessary. Lloyd is looking at using Net Cube as
a test package. Possibly Jay Martin can assist?

10. Alert Alert (Harry Lewis 4/18/97):
The description of the required data for a
alertRemovalOfBinaryChangeEntry was found on page 56 of the latest MIB
draft. It was agreed that this is not where this information is
expected and it should be moved or referenced. Randy will investigate
and propose a change.

Ron Bergman
4/23/97