Notes provided by Mark Needleman.

1. Status reports. Status reports were not given. Instead, status reports should be sent to the Maintenance Agency.

2. Register of Implementors. Ray circulated the register of implementers around for attendees to update and verify.

3. Breakout Meetings. Report on Breakout meetings held the day before the plenary:
   • Report on Bath profile  18 outstanding issues were discussed at breakout meeting  all 18 were resolved  list and resolutions will be posted for a couple of weeks for comments  goal is to get revised profile out in a couple of weeks

   • ZSQL breakout meeting reported on  2 items to be discussed at plenary

   • Report on Holdings Attribute breakout meeting  section in attribute set that related to old attribute architecture was deleted

   • Z39.50 and the Web  Ray will put out detailed report on meeting  decided not to focus on whether Z39.50 was in the client or in a gateway  issue raised on a directory of z39.50 servers  a idea was kicked around of doing a distributed service: in init response a server could provide a list of other servers it knew about  this would allow for distributed discovery of servers. XML and GRS issues were discussed. Decided to do an experimental encoding of the holdings schema in XML to get some experience and a better understanding of how XML worked  GRS will not be deprecated but XML will be introduced into more usage  For instance the schema for server information for distributed discovery will be done in XML  Also an analysis will be done of GRS and XML to see if an how the functionality of GRS could be done in XML- Ray will organize a subgroup to do this analysis. Joe Zeeman will lead the work to create the XML DTD for the Holdings Schema. There was some discussion of what impact having a DTD for Holdings would have on implementations of the schema using GRS. Discussion was held on what the ZIG should be doing about XML and RDF query work going on in W3C. Mark Needleman and Ray will server as ZIG liaison to XML Query  Sebastian Hammer was encouraged to serve as a liaison to RDF Query work

   • Ray will post a pointer to information on what XML schemas are and the functionality they provide. Andrew Goodchild will send Ray a pointer to a paper he wrote on XML Schemas.

   • Bib2 Breakout  progress was made on some small matters  question of using bib attributes to search authorities was discussed  Lennie will be looking at attributes from Zthes profile to support this  Lennie discussed work to move review of Bib2 out into the bibliographic community so as to get feedback from the relevant communities  she solicited input from the ZIG for people and listservs to send the Bib2 document for review.

4. Maintenance Agency Pages. Ray described and demoed the redesign of the Maintenance Agency Web pages. He solicited feedback from the ZIG on them. Entire text of Z39.50 is now available in HTML (as well as PDF). HTML version has embedded links. Members of the ZIG provided feedback on new design.

5. Re-affirmation. Z39.50-1995 is 5 years old and thus is up for re-affirmation in 2000. Ray outlined a re-affirmation plan (though this hasn't yet been coordinated with NIS0). A draft-for-ballot would be developed which would incorporate defect reports and amendments (those that have been approved by the ZIG); clarifications, commentaries, and implementor agreements will be considered case-by-case as to whether they will be incorporated. Some of the object definitions that are in the 1995 version will not be included, for example, the bib-1 attribute set and the bib-1 diagnostic set. These will be considered case-by-case. Ray suggested that the non-normative appendices not be included, however the ZIG wants them left in. Otherwise, there will be no new features and this will not be "version 4", however, work on "version 4" will continue in parallel. Nevertheless, it may be necessary to ballot this as a "revision" rather than a "re-affirmation". If the ballot passes, a single amendment to ISO 23950 could be developed, to keep these in synch.

6. Version 4. The issue of what to go in Version 4 was discussed  Ray solicited input from the ZIG on what should be in a future version of Z39.50 when the time came to develop one. There was some discussion of what version of ASN.1 to use in such a future version or whether to migrate to some other description language all together  there was further discussion of such things a hierarchical query, negotiating data models, and possibly adding in some features that were discussed during the development of the attribute architecture that couldnt be done within the constraints of the current protocol (these were mostly limits of the type 1 query)

7. Proposed Bib-1 Diagnostics. 2 new diagnostics for encapsulation and 1 for "no syntaxes available" were approved

8. Proposed ZIG Commentaries
   • Firewalls  approved
   • Presentation Contexts and Diagnostics and Diagnostic 106 - withdrawn
   • Server Default Record Syntax  withdrawn
   • The above 2 withdrawn commentaries will be replaced with a new Clarification that discusses the issues of preferred record syntax supplied and the case when preferred record syntax was not supplied  Clarification will be put out for a 1 month ZIG review and depending on outcome of that review may need to be discussed at the next ZIG meeting.

9. Negotiating Unicode and UTF-8. The 3^rd alternative, which depended on the change to character set negotiation, was approved.

10. Proposed amendment to Character set and Language Negotiation. Amendment makes collections OID optional and defines some semantics for when it's absent; adopted.

11. Proposed Clarifications
    • Multi-word Term with Structure word  if a multi word term is supplied with structure of word should be failed with diagnostic 108 malformed query  language in clarification referring to Bib-1 semantics document was removed.

12. Proposed Amendment for New Query Types. 2 alternatives were presented 1) to define a query type 104 as an external and then register new query types as OIDs using the external or 2) relax the requirement and allow new query types to be registered by the Maintenance Agency outside the standard  decision was to adopt query type 104 as an external and add an option bit to negotiate support for it. A new object class will be assigned for queries.

13. ZSQL extensions. Ray will assign OIDs as required.

14. Respective roles of GRS-1 and XML. Was discussed at breakout meeting and an analysis will be done

15. Future meetings
    • Leuven, Belgium  July 12-14 2000
    • Washington DC  December 6-8 2000
    • United Kingdom  May 2001 (tentative)
    • Rome  October 2001 (tentative)

    Format of current awareness/briefing meeting, breakout meetings and a final plenary will be continued with at least for the next couple of meetings  attempt will be made to more broadly publicize the briefing meeting which will be 1 full day  1 full day of breakout meetings, a half day plenary with time allowed for more breakout meetings after the plenary

    Some adjustment to this will be made for the Washington meeting due to availability of rooms.