Minutes of the Job MIB meeting in Boulder, CO October 31, 1997

Attendees

Ron Bergman - DataProducts
Carlos Cantu - IBM
Dennis Carney - IBM
Lee Farrell - Canon
Jeff Haemer - QMS
Tom Hastings - Xerox
Scott Isaacson - Novell
David Kellerman - Northlake Software
Harry Lewis - IBM Printing Systems
Jay Martin - Underscore
Kathy Melton - IBM
Ryan Nguyen - IBM
Jerry Podojil - Genicom
Stuart Rowley - Kyocera
Yuyi Sacchi - Japan Computer Industry
Jamsie Treppendahl - IBM

1. Ron Bergman (chair) reviewed editorial comments from the 10/27 
   e-mail - all were  accepted

2. A list of recommendations, submitted by Tom Hastings was reviewed.
   - Tom recommended adding Job-state-message natural language attribute
     * Accepted
   - Tom recommended adding the choice of  octet string vs.   
     displaystring for jobCodedCharSet so either the IANA 
     register printer MIB values can be used or the IANA MIME 
     types can be used.
     *Dave Kellerman indicated choice would over unnecessarily burden             the client application. Harry Lewis would like to stay with enums             if only 1 is chosen. Topic to be further discussed on mailing       list.
   - Tom recommended adding a natural language attribute for the job.
     *Tom will write up exact text and circulate on the mailing list. 

3. Standards track vs. Informational. 

There is still consternation over the fact that the Job MIB charter never seems to have evolved under the IETF. There is a view that our efforts should have been directed by the Management rather than Applications area, in which case the IETF may have looked more favorably on our MIB. It is noted that the Printer MIB may actually be more widely deployed than any other MIB in the IETF! (many printers per one router). Part of the discussion related to the fate of the Printer MIB, also. 

Randy Turner indicated, during an earlier PMP discussion, that he would still like to see the Printer and Job MIBs become IETF standards and feels that Experimental is a more appropriate and fruitful route than informational. We need to get in touch with the Area Directors of both Management and Applications and have an IETF decision made. Two week deadline to resolve else submit as informational RFC. Also, find out what area is guiding the hrMIB, today, and who is heading that project.

If we go the Informational route we will register the Job MIB under the new PWG OID as follows:

 PWG OID = ... Private.Enterprises.1699 so the Job MIB will be
           ... Private.Enterprises.1699.1. The first assignment 
                                           of the job mib will be
           ... Private Enterprises.1699.1.1

4. Mapping document. 

   - LPR/LPD - Header of data file contains host name. Typically does 
     get passed to controller. Used to create jmJobSubmissionID format 
     9. (See Ron's proposal). In peer-to-peer environments, the host 
     name is typically derived from the client workstation. In 
     client/server the host name is usually that of the server.

   - We decided to make jmJobSubmissionID format 9 ONLY apply to LPR

   - We added format A for Apple Talk.

   - The IPP jmJobSubmissionID format is derived directly from the 
     printer URL. There are no other alternatives. By definition,
     because IPP is bidirectional, it benefits from having the                jmJobIndex returned via IPP during submission. 

   - IPDS - Harry will investigate possible mapping or changes to IPDS 
     to support the Job MIB

   - NPDS - Provides jmJobSubmissionID. Does not map to jmJobIndex.

   - PJL -  Implementation note. JobID table entry is likely to be 
     created when the up front PJL has been parsed in well behaved, 
     coordinated job submission systems.