"Future of Z39.50"

ZIG Meeting Breakout Session

July 12, 2000

Discussion Points

1. "Maintenance" Revision

Will incorporate defect reports and amendments (those already approved) as well as commentaries and implementor agreements as appropriate. Each will be considered individually.

Appendices:

• Appendix 1 (object identifiers): keep.
• Appendix 2 (application contexts) omit.
• Appendix 3 (attribute sets) the list of sets at the beginning of the appendix will be dropped. The text immediately following will be integrated into the body of the standard. Bib-1 will be dropped from the appendix (it will be maintained solely on the Maintenance Agency web site). Attribute sets exp-1 and ext-1 will be retained in the appendix.
• Appendix 4 (diagnostics) the bib-1 diagnostic set will be re-named (proposed name: general diagnostic set). This set will be retained in the text and brought up-to-date. Diag-1 will be dropped from the text (as far as we can tell, nobody has implemented it).
• Appendix 5 (record Syntaxes): the following definitions will be retained in the text: SUTRS, GRS-1, Explain, and Extended Service. (All other information will continue to be maintained by the Maintenance Agency.)
• Appendix 6 (resource report formats): Background: Orig0inally there was resource-1, which listed 16 categories of resources, with no provision of extensibility. Resource-2 was added in 1995, inheriting the original 16 categories, with provision for extensibility. Thus Resource-2 is a compatible superset of resource-1. We propose to drop resource-1 and retain resource-2 (with comment to the effect that a client might want to be prepared to recognize the resource-1 object identifier.)
• Appendix 7 (Access Control formats): to be retained in its entirety.
• Appendix 8 (Extended Services): to be retained in its entirety.
• Appendix 9 (User Information Formats): this appendix is divided into sections USR.1 and USR.2; USR.1 is searchResult-1, which will be retained. USR.2 discusses negotiation, and this text is superceded by the negotiation model. We propose to drop this section, and add a new appendix covering negotiation.
• Appendix 10 (element specification formats): We propose to include the eSpec-2 definition, with a note about eSpec-1, including a a comment to the effect that a server might want to be prepared to recognize the eSpec-1 object identifier. We propose to include eSpec-q.
• Appendix 11 (variants): Will be retained and brought up-to-date.
• Appendix 12 (tag sets): Wwill be retained and brought up-to-date; semantics will be integrated into the text. The list of schemas will be dropped. Appendix 13, 14, and 16 (non-normative): retained.
• Appendix 15 (profiles): dropped. A non-normative appendix is proposed, discussing profiles, how they are developed, and their relationship to the standard.
• New appendix: Attribute architecture (non-normative).

A draft for discussion at the December ZIG will be prepared, and a draft for NISO ballot may be ready during first quarter 2001.

2. Z39.50 "lite".

Is this still a viable idea?

3. ASN.1 or XML?

There appears to be general agreement that going from BER encoding to XER is mechanical and should be left to implementers. Replacing the ASN.1 description with an XML description is far less clear. Though there seems to be immediate demand, would it be a good idea, to try to understand the magnitude of the effort, to commission someone to write an XML description of the ASN.1 PDUs.

4. Call to Implementors

There will be a call to registered implementers and other relevant parties, to explain that Z39.50 is up for 5 year review and that a "maintenance revision" will be recommended, but that we would like to start planning for future directions for the standard and are soliciting comments and suggestions: what would you like it to do, what are the problems? Should there be a "version 4"? What direction should it take? Would it have to be compatible with v3? with v2? If there is much response, there may be a follow-up workshop, using the responses as a guide: who to invite and how to structure it.

5. Long term future

Start thinking about the functional requirements for intersystem searching/information retrieval in the next decade. Where does Z39.50 fit (if at all)? May start to plan for a workshop with invited papers.

6. National profile and holdings schema

7. Role of the ZIG in Z39.50's Future