Attending:Marianne Burke, Bob Buckwalter, John Collins, Michael Fitzgerald, Dale Flecker, Rodney Goins, Lynne Schmelz (chair), Ted Pappadopoulos, Jon Rothman, Dorothy Solbrig, Judy Warnement, Robin Wendler (recording), Charles Willard
Absent: Carrie Kent, Mark Van Baalen
Guests: Diane Garner
Jon Rothman followed up on a memo from Heather Reid. The GIRLS committee reports which HAAC reviewed last year recommended a fundamentally network-based service, with access to be provided from network attached devices and some lower level of service to be available from terminals. In the intervening months, the working group has concluded that the Bridge cannot be located on the mainframe because the mainframe is a poor network client, and that terminals connected to a network-based (i.e., non-mainframe) Bridge would not function well enough to warrant the resources necessary to design and support it. The group recommends that access to the Bridge be available from the following devices:
Hard-wired terminals on traditional mux-modem connections would not be able to access the Bridge.
Dale Flecker pointed out that historically HUL had maximized investment in all devices, basing every development on the common denominator. With this project, in order to create a reasonable service we are forced to go in a new direction. Charles Willard asked whether libraries which would lose Bridge access had been represented in the decision process, and was assured that they were. He expressed concern that we would be making unfortunate distinctions, and that this change in thinking had not been well-communicated. Dale explained that this is not a bad direction to be moving in; rather, it is inevitable. The next generation of library system will almost certainly run only on networked devices. Libraries should always be on a rotating equipment/infrastructure upgrade schedule. Only the timing of the shift is unfortunate in this case. He added that this was not a quick or easy decision. Nine months of careful evaluation led the group reluctantly to this choice.
Other HAAC members expressed support for the recommendation, and HAAC unanimously approved in principle abandoning plans for universal terminal access and proceeding with investigation of gopher technology.
RECON is going extremely well. The claiming and keying services are ahead of projected targets for first year production. OCLC is converting 75,000 records/month now, and expects to have excess capacity beginning in January. OCLC asked if Harvard wanted to speed up production, and after evaluation of technical concerns (including frequency of HU searches retrieving the maximum number of hits, HOLLIS response time, effectiveness of duplicate detection) and consultation with the libraries, HUL agreed to increase the rate of recon by a third, up to 100,000 records/month. The queue will be reworked in January to reflect the compressed schedule. At the increased rate, RECON may be completed by the end of 1996.
The heading correction process has been completed on the base file (HU records through September 1992). Approximately 800,000 field-level corrections were made to the approximately 3 million bibliographic records sent. Nearly 500,000 authority records have been received and loaded.
Diane Garner explained the need to get receipt and cataloging of government documents under more efficient control, including subscribing to some form of electronic shipping list service and automatic receipt of cataloging for profiled Items. Her goal is to get prospective GPO tapes loaded during this fiscal year or early next fiscal year. She envisions a two phase process, with current receipts handled first, and retrospective coverage obtained in the second phase. Documents are most valuable when new, which adds to the need to provide easy, timely access.
There are a number of options for receiving both shipping list data and cataloging data. The records for the two stages, receipt and cataloging, may be used together in one system or reside in separate systems and be used to control separate parts of the process. Two options were discussed in depth.
The first option is to use a stand-alone check-in system designed by the University of California at Riverside, which uses a PC-based database manager to manipulate shipping list records for receipt. We would have to profile separately to receive cataloging records for loading into the HU file in HOLLIS, and design and program some mechanism to compare the PC database to the HU database in order to determine which received items had never been cataloged and which items were cataloged which had never been received. This option uses existing services and allows for a high degree of local control over check-in information. Its disadvantage is that the information for items which had not yet been cataloged would be available only on a PC in the Government Documents division, and it would be cumbersome to identify discrepancies between received items and cataloged items.
Another option would convert shipping list records into MARC format for loading into the HU database. Staff could check in on these records. With this model, as with the first model, profiled cataloging records would be loaded into HU as they became available. In this model, however, they would merge with the check-in record already in HOLLIS. Bernan/Unipub and Peter Ward have been negotiating an arrangement whereby Bernan would supply profiled shipping list data, including a Bernan control number, to Ward, who would FTP it to Harvard. The records could be loaded into HU with location fields coded to suppress the record from public display. As staff received items, they could change each record to display. Ward would match these check-in records against GPO cataloging and when a match was found, add the Bernan control number to the cataloging record and FTP it to Harvard. HOLLIS would match the cataloging with the check-in record using the Bernan control number. An advantage to this method is that the status of uncataloged items would be available across campus in HOLLIS, and the Law Library could also participate.
There was much discussion. One of the concerns in evaluating any option is the timeliness of record receipt. In order to be useful for check-in staff, the shipping list records must be available in some local system when the order is received. Another necessity is the provision of detailed Harvard location information in the shipping list record, so that staff members can determine where to send an item by looking at the shipping list record, and not be required to look up an Item number/location table in yet another file.
There was agreement that the Bernan/Ward option was desirable if they could provide the necessary level of location information in the shipping list records. If not, Diane would prefer the check-in capability of the standalone system. Harvard representatives will meet with Bernan/Ward at ALA to determine whether their service will fulfill our needs, and Diane will come back to HAAC with a more concrete recommendation.
Dorothy Solbrig listed topics which SSHUSH had addressed this year, including consideration of indexing needs for physical form/genre information (655/755 fields). A joint workgroup was formed with Bib Standards, and a recommendation is due by December 1. SSHUSH also wrote and reviewed specifications for the course reserves project, learning much about the structure of the new database (which differs from HU in that it must display course and reserve information) and the varying needs of different reserves units. New fields for course information and specifications for indexing them were approved. New display tables may be needed. SSHUSH wants subject headings to display in the long display only in the reserves database, and wants LOCATIONS: to display as REQUESTED AT: since some materials may not ever actually be at the library.
Michael Fitzgerald noted that Bib Standards has proposed a number of indexing and retrieval enhancements during the last year. The group also agreed that authority record functionality be extended to alternate institutions (specifically the Physics Preprint database). Michael noted that turnover in OIS had slowed the group's work somewhat. Michael and Dorothy are working with Jon Rothman from OIS to screen and direct issues to SSHUSH and Bib Standards, forming joint task groups as appropriate, in order to make the operations of both subcommittees more efficient.
The minutes of the previous meeting were approved.