From: Bill Cattey Sara Randal's machine containing the minutes of the previous ZIT has a cold. I have been asked to send what I had taken as a preliminary version to server as a memory jogger for the upcoming ZIT. Please forgive inaccuraccies, omissions and / or incomprehensible abbreviations. Respectfully submitted, -wdc -----PRELIMINARY----- Attendees and what they're doing: Jim Michael -- Data Research. Using ISODE, and just received patch from digital for the recursion problem. Have clients up. Working on server now. Peter Ryall -- Mead Data Central. Solidified design. Phased. Skeleton in june more full in 3rd quarter. Using ISODE ASN.1 Ralph Orlik -- Library of congress Robert Waldstern AT&T -- Up and running. 2 Databases: book database with large records. Libnet statistical database -- Has no formatting. Server is up 24 hrs a day. John Kunze -- Berkeley. Client and server. Up 24hours a day (version 2.2) . Good reference implementation for demonstrating ASN -BER/TCP. Source code release scheduled for 1st week April (balloted version). Infocal.berkeley.edu. Presenting paper on non-bibleographic applications. Bo Parker -- Stanford. Client for Folio(?). Still finding right boundryfor how much of the mainframe software to use. Harold Finkbeinder -- Stanford. Doing the work. Sarah Randal -- NOTIS -- both client and server. Looking at ASN1 compiler stuff now. Bob Shafer -- CARL systems. Added staff. Learning. Identifying tools. Looked at the PL/1 stuff. Ron Larson U. of Maryland -- Working with CARL Harry Morris - TMI-- Wais research center of UNC working on '92 protocol for WAIS. Bill Cattey -- MIT -- Learning. Now aware of several potential partners. Looking to have demonstration system up within 12 months containing the MIT Library catalogue, a search engine, and clients tied together with Z39.50. Janet Vratney-Watts -- Apple Kazu Yanagihara -- Apple -- looking for a partner. Has few resources. Lennie Stovel -- RLG -- Working on server as most important focus. Took an old LC client for testing. Making slow progress. Building a new search service based on Z39.50. Ralph Orlik -- Library of Congress. Switching to TCP. Emphasizing client first and expect it will be ~9 months before they're interoperating. Cecilia Preston -- UC Note taker. Clifford Lynch -- UC moderator. Eric Ferrin -- PennState Mark Needleman -- UC Steve Silberstein -- innovative. Just getting started. Hope to get a client up soon leveraging off other partners. Michael Thwaites -- UC DLA. Making progress. Server up 24 hours a day. DLA.UCOP.EDU. Mark Kibbey -- -- Have proprietary X Motif client that they want to migrate over to Z39.50. ?? -- RLG The discussions: Clifford: Balloting going along OK. It doesn't seem like there will be big changes. Error ADPU: Dropping it for now. Forwarding to ZIG. Instead we'll use Resource Control, if anything. ---- Briefbib: Change the name to 'OBJECTSUMMARY'. John Kunze's extensible system is relevant. It could provide a mechanism for implementing briefbib. Issue: Can you implement a simple non-extensible subset of John's stuff to get briefbib through his syntax? Yes. Issue: Briefbib was originally intended as a least-common-denominator use set. Issue: Some sites have only B IB-1,and adding a second record syntax is a non-trivial exercise. These people would prefer to send a full BIB-1 with lots of the fields left blank. Add another optional field that describes what kind of document is being described. (Abstract, book etc.) Change size to 'visible optional' Get a registry number for this from the ZIG. Expect that it will become obsolete in the future if John's extensible system is adopted. All servers should not choke on the 'B' atrribute. ---- OPAC Marc: Get a registry number for this from the ZIG, and adopt. ---- Interoperability agreements -- amendments and discussion: 2. Don't need to support present. 3. Strike 'user information field' and revise: We will try not to send fields longer than 64 bytes. 4. substitute 'accept' for 'support'. Add in wording saying we send ANSI to each other for the time being and eventually will send ISO identifiers. 5. Support transfer syntaxes OBJECTSUMMARY, and/or USMARC, or other appropriate syntaxes. The objective is to have everyone support either OBJECTSUMMARY or Info-1 as a common transfer syntax. Consolidate 6 & 10. 7. Support type-1 queries. All others are optional. 8. dropped. Any error text sent is encoded in a BER encoded ADPU. 9. later discussion 11. add 'B' element support. add: 12. Recognize ASN.1 changes agreed upon by ZIT ---- ASN.1 changes: All agree to accept PDUs which contain additional information field which is defined as type ANY. Field number 210. ---- Can we recognize 'xxdefault' case insensitive as a default database name. ---- TCP/IP layering review: Resource report stuff is a potentially serious problem. Will take it up with the ZIG. Can we make the Resource Report to be an External? Does doing so represent a backtrack from work that the ZIG did? No --we just register the present definition as part of the BIB-1 attribute set. ---- Additional interchange formats: Few people plan to need Kunze's extensions in the next 6 to 8 months, but many expect to in the 12 to 18 month timeframe. John will implement his example version and seek feedback from the ZIG and interested parties here in the ZIT. ---- Next Meeting: Tacked onto the day after the Chicago ZIG, June 4. ---- Status of explain: A fair number of people expect to implement it or want to use it as clients in the next 6 months. ---- Bob's extra info. ---- Demonstrations: Demos are important. We should get some going by the fall. They require more preparation than many people believe. Clifford will demo at Net'92. Let's target LITA, Sept 13-16th, for some sort of demo there. Clifford will ask the LITA people (and the TESLA people who apparently have a Z39.50 demo scheduled.) if there's some central place appropriate for us to have an interoperability demo. EDUCOM vs. INTEROP: On opposite coasts the same week. Let's pursue EDUCOM. 28-30 October. It has the same vendors as INTEROP but a clientele more closely related to our membership. CAUSE is a possiblity, but let's see how the first 2 demos work out. ---- Of ZITs and ZIGs: Mark Hinnenbush, Ray Dannenberg, and Clifford will come up with a document which will restate the goals of the two groups, and the charter of the ZIT. Our focus is on the short term and on interoperability, whereas the ZIG is focused more upon protocol issues. Operative rule of thumb: Any problem on a timescale farther in the future than 6 months is a ZIG issue, not a ZIT issue.