PWG MIBs MEETING MINUTES
Savannah, Georgia
October 2, 1998


Job MIB

Although the Job MIB has been a completed PWG standard  for several months, there have 
been some recent comments from the IETF in response to our efforts to have them published 
as RFC standards.

1. The MIB is entirely R/O. (This is by design).
ACTION:  Respond to David Perkins explaining that this is only a monitoring MIB.  For 
security reasons, the Working Group agreed that the MIB would not contain any read-
write objects.

2. We have inadvertently misused the REFERENCE clause. 
ACTION:  Delete all REFERENCE clauses and move reference information into the 
DESCRIPTION clause when appropriate.

3. We need to move the definition of enumerated labels to a new section outside
of the MIB portion.

4. We MIGHT have  used too many SHOULDs and SHALLs (surprised? ;-).
ACTION:  Review the use of these words and remove or change to lower case any that are 
not absolutely necessary.

5. JobProcessAfterDateAndTime - there was a comment about improper or
nonstandard use, but we do not understand the comment. We need to clarify with
the reviewer.
ACTION:  Respond to Keith McCloghrie explaining that this is not a standardized display 
format.

Tom Hastings comments:
1. A new draft was presented with changes in response to items 2, 3, and 4.
2. Two new attributes added - mediumTypeConsumed and SizeConsumed.  Both 
attributes were previously approved as additions to the MIB.
3. Tom will remove attributes from the table of contents just for brevity.

New Issue:
Tom Hastings presented a problem discovered by Xerox in using the GetBulk 
operation with the Attribute table.  The present table configuration allows GetBulk 
to work to obtain all the attributes for a given job or a number of jobs.  Tom has 
proposed a mirror table that inverts the order of the indices so that a GetBulk can 
obtain an attribute for all jobs.  This addition would result in enhanced performance 
on the client side.  Tom will draft the required text and post to the mail list.

Today - the Attribute Table is indexed by:
 	-  Job index
 	-  Attribute type index
 	-  Attribute instance

The proposed Mirror Table would be indexed by:
 	-  Attribute instance
 	-  Attribute type index
 	-  Job index


Job MIB interoperability testing may be possibly within appx. 6 months.



Finisher MIB:

MIB Changes:
1.  Moved the MODULE-IDENTITY to  { experimental 64 }  Accepted
2.  Added the following enums to FinStitchingTypeTC:   Accepted
         stapleDualTop(10),
         stapleDualBottom(11),
         stapleDualLeft(12),
         stapleDualRight(13)
3.  Added a new attribute and textual convention:   Accepted but change to a type 2 enum 
and change bottomUp(5) to bottomUp(4).
         stitchingDirection(31),            StitchingDirTypeTC
            INTEGER:  Defines the orientation of the stitching
              process.

StitchingDirTypeTC ::= TEXTUAL-CONVENTION
-- This is a type 1 enumeration.
    STATUS       current
    DESCRIPTION
      "Defines the direction, relative to the top sheet in the output
       subunit, that the stitching operation was performed.  For a
       topDown(3) process, the staple will be clinched on the bottom
       of the stack.  This parameter can be used to determine what
       order the pages of a booklet are to be printed such that the
       staple clinch will be on the inside of the resulting booklet."
    SYNTAX       INTEGER {
         unknown(2),
         topDown(3),
         bottomUp(5)
        }

4. Removed the following note from section 6.1:   Rejected, new enum is changed to type 2.
	"There are no type 1 enums in the current draft."
5. As a result of the discussions concerning the Job MIB it was also discussed and agreed to 
move the specification of the attributes used by FinAttributeTypeTC out from the MIB 
body and into the text.
New Issues from Sharp:
1. Several index objects allow a value of 0.  The minimum value should be 1.  The following 
index objects will be changed to correct this problem.
      finDeviceIndex
      finSupplyIndex
      finSupplyMediaInputIndex
      finDeviceAttributeInstanceIndex

2.  There is no enumeration in prtMarkerSupplyUnitTC that is appropriate to staples.  It was 
agreed that an enum, such as "units" or "each" should be added to the TC in the Printer MIB.


Common Finisher Interface Discussion:

Kevin Palmer (Duplo) had agreed to generate a proposal for the Finisher Device Command 
Set and the Finisher features, functions, characteristics, and status.  Kevin was unable to 
attend the meeting, but he did submit a list of items that pertain to the above subjects.  After 
a brief review, the group decided that there was not sufficient information to generate 
further discussion.  No additional work on this topic will occur until Kevin's complete 
proposal is received.


Harry Lewis - IBM Printing Systems
Ron Bergman - Dataproducts Corp.