Agenda for HOLLIS Liaisons Meeting #83
January 13, 1993
Lamont Forum Room, 2:30 - 4:00 p.m.
1. Announcements: Tracey Robinson
Spine labels from HOLLIS coming soon. The long awaited HOLLIS spine label function is in the final stages of development; OIS anticipates it will be available to library staff in early 1993. Further information about this function will appear in the February 1993 issue of this newsletter.
MeSH in HOLLIS. OIS plans to add National Library of Medicine subject headings (MeSH) to HOLLIS. Development is already underway and OIS anticipates that MeSH headings will be added as a technical services database to the SH database sometime this winter. More information will be forthcoming as development proceeds.
Proposed demise of HOLLIS EMS. OIS is investigating alternatives to the current EMS software available within HULPR. EMS provides limited mailing capabilities and a bulletin board for HOLLIS staff operators. The overhead and cost of maintaining EMS is becoming a burden not justified by the software's usefulness. Since the bulletin board is the most used feature of EMS, one alternative would be for OIS to develop a similar bulletin board function.
OIS is seeking comments from EMS users about removing EMS from HULPR. Please contact Tracey Robinson with your comments.
Payment flagging update. Due to a programming glitch, some pay statements last updated between 7/1/90 and 12/31/91 were not successfully flagged during the last payment flagging cycle. OIS will run the cycle a second time to catch the missed statements, and will reissue the cumulative payment statement microfiche (from 7/1/89 to 12/31/91). [Editor's note: the second cycle of pay statement flagging ran successfully and the replacement microfiche has been distributed.]
There has been some confusion about the disposition of the HOLLIS microfiche that is produced as part of the various HOLLIS weeding cycles. Some fiche are produced annually in July, representing fiscal year data, while other fiche are produced whenever acquisition data is weeded from the online file. Attached to this newsletter is a complete list of the fiche which should be kept IF you choose to maintain a complete inventory of HOLLIS microfiche. Libraries are not required to maintain full backlogs of the HOLLIS microfiche but should maintain whatever fiche may be needed to resolve acquisitions problems which cannot be resolved by data maintained online. The development of appropriate retention policies are the responsibility of each HOLLIS participating library. Contact Tracey Robinson in OIS if you have questions.
HOLLIS network issues. OIS Network Support reports that the majority of new installations and HOLLIS network problems have been satisfied. Tracey noted that for installation of new terminals (not replacements of existing terminals) there are few unused mux ports available, so OIS has been doing some juggling -- one more reason for units to contact Ciata Victor in OIS as soon as they begin considering new installations.
A list of outstanding network service requests requiring attention appears in the Network Notes section of this newsletter. If you have a problem or plan an equipment installation that is not on this list, please contact Ciata Victor in OIS.
VINE update. The Veritas Information NEtwork (VINE) -- Harvard's new campus information system -- is scheduled for a "soft launch" in mid-December. [Editor's note: the VINE was successfully launched "softly" on December 18.] This "soft launch" included the availability of VINE via the Internet, ISDN dialup, and on library HOLLIS terminals at Gutman Library. Shortly, VINE will hopefully also be available from the HOLLIS public terminal in University Archives. The official debut of the VINE is scheduled for the beginning of the spring semester, when VINE will become available at HOLLIS terminals in most Harvard libraries. To accompany the official release will be public information sheets describing VINE and signs for HOLLIS terminals on which VINE becomes available. OIS is currently investigating analog dialup access to VINE and it is unclear whether this access will be available with the spring debut.
In the interim, staff interested in taking VINE for a "test drive" can access it several ways:
Tracey reminded liaisons that some units had not yet returned the VINE form on which a library specifies which HOLLIS terminals will offer VINE as a service. Please fill out and return these as soon as possible. As the "hard launch" approaches, the VINE Office will provide more information about terminal signs and information sheets. Units with questions about VINE can contact Project Manager Amy Lozano at 495-1821 or e-mail her at LOZAN@HARVARDA.
Flow of Recon records into HU increases. Karen Carlson Young, Recon Project Manager, reported on the latest Recon developments. OCLC processed approximately 13,000 Widener/Houghton records in November, of which about 9,000 were added to HU on the November OCLC tape. So far, the hit rate for claimed records is 68 percent, slightly higher than Harvard's original projections. Starting in January 1993, searching will reach approximately 50,000 records per month when the Design Library enters the Recon queue. Gutman and Law Library are scheduled to begin in May.
Now that the flow of records into HU has increased, staff may notice these records and have questions about their contents. Keyed Recon records have an encoding level of "K" (less than full [OCLC member]) in HOLLIS. One notable aspect of these records is that their subject headings will most likely be coded as local (that is, stored in a 69x field in the HOLLIS MARC record). Local subject headings are only searchable by keyword in the Public Catalog. Consult an article in this issue for further details about treatment of Recon subject headings in HOLLIS.
The Recon Office is about to begin quality control sampling of Recon records in HU. Maria Quintana from the Widener Cataloging Services Department will join the Recon Project to work part-time in the quality control process. If you have questions about Recon, please contact Karen Carlson Young at 496-4011.
Update on electronic transfer of payments (ETP). During the first two weeks of electronic transfer, 1,200 corporate invoices were approved, and 55 percent of these were processed using ETP. OIS estimates that approximately 60 to 70 percent of Law (department 57) and HCL (department 58) invoices will be processed through ETP. [Editor's note: Since the December liaisons meeting, the Kennedy School Library has begun using ETP as well.] There was a bug in the ETP process that caused the HOLLIS invoice control number (ICN) to print on check stubs instead of the vendor's invoice number, but Daniel Bednarek reports that the problem is resolved. Units suffering from this bug will receive reports of the affected invoices. If you have questions about ETP or are considering participation, contact Daniel Bednarek in OIS.
Morse music scores in HOLLIS. Suzanne Kemple reported that the conversion into HOLLIS of records for Hilles' Morse Music collection is now complete. The approximately 5,200 bibliographic records representing this collection are now available in the HU database.
Tracey Robinson summarized for liaisons some of the highlights from the Fall Managers' Meeting:
OIS update report. Dale Flecker updated managers on OIS development activities during the previous year, including: Recon, Library of Congress Books File, global change function, electronic transfer of payments (ETP), course reserves function, various telecommunications projects, VINE, online course catalogs, and the Fall 1992 release of the HOLLIS Public Catalog (Anthropological Literature, Religion Index, and several display changes).
Business Resumption Project update. Tracey Robinson updated managers on the Business Resumption project (also known as "disaster recovery") which got underway early in 1991. Since that time, OIT has signed a contract with a vendor to provide "hot site" recovery service in the event of a major disaster at the OIT data center. The "hot site" is a mainframe computer which would support HOLLIS in an emergency along with other critical Harvard University applications. OIT and the University Library are beginning to develop detailed recovery plans . OIS expects that a test of the recovery plan will take place by the end of this fiscal year.
Future HOLLIS network options. Heather Reid presented a detailed description of the network development project currently underway in OIS. In concert with Business Resumption, the goal of this project is to provide a more flexible way to connect terminals and microcomputers to the HOLLIS mainframe computer. The approximately 500 IBM 3151 and 3163 terminals in libraries are connected to the OIT mainframe via telephone lines (LAD circuits). In the future, they will connect via Harvard's High Speed Data Network (HSDN). Eric Young in OIS can provide copies of a paper written by Heather Reid which discusses these connections to the HSDN. OIS expects a slow migration of IBM terminals to the HSDN over the next few years.
Library staff using the HULPR system will also benefit from network developments to allow HULPR access from microcomputers over the HSDN. The Botany Libraries are currently testing this connection method and OIS will keep everyone informed as development proceeds.
Managers also discussed alternate billing schemes for network-connected devices using HULPR. There is currently a fixed monthly rate for all dedicated HOLLIS/HULPR terminals. Many HOLLIS managers noted that the level of activity which one might expect from a microcomputer user in HULPR would be significantly less than the activity level expected on a dedicated "dumb terminal". This argues for a different rate or method of charging for HOLLIS activity for network attached microcomputers. OIS is investigating providing a fair billing rate for these new types of HULPR workstations.
Distributable Union Catalog (DUC). The DUC, the microfiche backup for the HOLLIS Public Catalog, has grown so expensive that most Harvard libraries no longer can justify their subscriptions. Recon will add to the DUC's growth rate, putting even a single copy of the DUC out of the reach of most libraries. HOLLIS Liaisons and SSHUSH have discussed the possible alternatives, which include: shrinking the size of the DUC in some way (for example, removing the subject headings section), moving to a CD-ROM product, and eliminating the DUC altogether. Some managers favor a plan to load a copy of the HU database into RLIN and use that as a backup in case the Public Catalog becomes unavailable. Harvard would send periodic updates of HU to keep RLIN relatively current. Other staff members question whether RLIN is an effective backup alternative. No firm decision has been made on this issue, but it is possible that the January 1994 DUC will be the last microfiche DUC produced.
HOLLIS reporting rates. OIS has sensed growing library dissatisfaction with the fee structure for HOLLIS bibliographic and statistical reports. Managers discussed alternative fee structures that might be more acceptable.
Currently, OIS charges $30 for small reports and produces larger reports at cost. Actually, a "small" report is one that costs OIS less than $100 to produce (computing and printing charges) -- any report with an actual cost of less than this amount will cost $30. Reports with an actual cost of more than $100 are produced at cost (i.e., the full cost of the report is passed through to the requesting library). In addition, there may be programming charges for reports of any size if a programmer spends more than 30 minutes producing a report. OIS programmer time costs $30 per half hour.
Managers considered a proposal to modify billing for fiscal 1994 that would reduce the cost for many reports. In the proposed new world of reporting, the fee structure would have more levels. Reports produced on a regular basis (e.g., monthly, quarterly) would cost $10 per run. Reports produced on an ad hoc (i.e., one-time) basis would cost $50. The $100 threshold would be eliminated, but if any single report exceeds 200 pages in length, OIS would pass the cost of printing onto the requesting library (OIT currently charges $.07 a page). Any report that requires more than 30 minutes of OIS programmer time would incur the traditional programming charge.
HOLLIS End-user reporting project. OIS is investigating the implementation of end user reporting in HOLLIS. Such a development would allow HOLLIS staff operators to generate their own HOLLIS reports using local library microcomputers and printers. Heather Reid from OIS heads this project, which should lead to a prototype system that could be tested in the spring of 1993. Stephen Stillwell asked whether OIS could supply HOLLIS reports on diskette rather than paper. Tracey replied that although there have been several trial reports produced on diskettes, this is not yet considered a fully supported and universally available product. Kate Ellis interjected that circulation notices and reports are now produced in machine- readable format for Countway and Cabot Library has received circulation reports on disk as well. Libraries interested in machine-readable reports should contact Kate Mullen in OIS.
Notes and Reminders
OIS and HOLLIS liaisons agreed on this guideline last year as a way to reduce the volume of payments in the flagging process and increase efficiency. Any units that maintain serial or work-in-part orders that are mistakenly coded with something other than SCOPE=2 will not have their pay statements flagged for weeding. Units can change SCOPE to "2" to insure that future payments are "flaggable." If you have further questions contact Linda Marean in OIS.
OIT produces "FAQ" sheet on HOLLIS dialup problems. OIT is about to introduce a new line of information sheets called FAQ Sheets (Frequently Asked Questions). The first such sheet provides troubleshooting instructions for users dialing up to the HOLLIS Public Catalog from a microcomputer. The staff of the OIT Hotline (known to us as user assistance, 495-9388) have put this sheet together using the common problems reported to them by users. OIT also has plans to produce a FAQ sheet on the HUBS system. As a service to libraries, OIS plans on distributing master copies of the dial-up FAQ sheet. OIT can supply copies as well. A copy of the dialup FAQ sheet is attached to the end of this newsletter. If you have questions about this sheet, please contact OIT user assistance at 495-9388.
HOLLIS/HULPR downtime on 16 December. HOLLIS operations were interrupted on Wednesday, 16 December, by two separate but related problems. Here is a play-by-play of events.
HOLLIS/HULPR downtime. OIT made some scheduled system changes in the early morning of 16 December. HOLLIS and HULPR came up normally. Around 8:00 am, the MVS operating system under which HOLLIS and HULPR operate went into a "hard loop," which is a bad thing. To interrupt the loop, OIT had to bring down all systems. OIT then had to back out the system changes they had made, which apparently had led to the loop in the first place. This "backing out" took most of the morning. All systems were operational by 11:30 am.
MITEK downtime. Late that same morning, the MITEK box (through which certain HOLLIS terminals and network PCs communicate with the mainframe) crashed, apparently as a result of the system problems described above. Most OIS terminals and PCs, along with some terminals/PCs in Widener and Countway were down most of the day. Actually, to be accurate, they were up and down, up and down, up and down ... The system problems suffered earlier in the day seem to have corrupted the software used to operate the MITEK. A backup of the software was installed and the MITEK is functioning normally again.
OIS and OIT are working to analyze the events of 16 December; we expect that this problem will not recur. Our sincerest apologies for any disruptions to your work day caused by this downtime. Contact Linda Marean in OIS if you have further questions.
AL database quarterly update. The first quarterly update of the Anthropological Literature database -- 2,275 new records -- was completed on 10 December. Although database updates do not normally appear in this newsletter, the Editor just wanted to remind HOLLIS users of the AL database update cycle and remind them to consult EMS common messages for update information on all HOLLIS databases.
If you use MARC field 752, read this... The 752 field (hierarchical place name) is used by a few units at Harvard to identify the place of publication of a book or the community served by a newspaper. Because of the recent changes in the former Soviet Union, the Judaica Division of the Harvard College Library needs to change a large number of 752 fields to reflect the new forms of the place names. The new Global Change facility in HOLLIS can be used to make these changes, however, they would affect not only those 752s in Judaica records, but all 752s which correspond to the obsolete form.
While there is no published authority for the form of name used in the 752 field, there is agreement that current forms should be used regardless of the form of the name appearing in the publication. I would like to know which units now use the 752 and whether they anticipate any problems arising from this proposed maintenance via global change. Please contact me (Robin Wendler) at OIS or via e-mail at RRKING@HARVARDA.HARVARD.EDU.
Response Time Testing. The early December timings reveal a noticeable increase for response times in the 4+ second range. The November tests recorded only 3 percent of searches in the 4+ second range while the December readings had 12 percent of responses in the 4+ second range. Otherwise, response times remain excellent. With the assistance of our response time volunteers, OIS will continue to closely monitor these timings.
Response time tests are scheduled for the first Tuesday of each month. Upcoming tests are scheduled for: 5 January, 2 February, 2 March, and 6 April. Volunteers should mark their calendars appropriately.
Recon subject headings. The treatment of subject headings on recon records will vary depending on whether the bibliographic record was claimed or keyed and may vary from library to library, depending on its particular cataloging practices. Basically there are two aspects of subject headings that must be considered for both current cataloging and for recon records. The first is whether a library uses LC or NLM subject headings and the second is the tagging of the subject headings as local or national.
For current cataloging the set of subject headings on incoming records that is kept depends on the library's profile. If the library is profiled to keep MeSH headings, then LC headings on an incoming record for that library will not be retained. A library which is profiled to receive only LC subject headings will not have MeSH headings retained on any of its incoming records.
For recon records both sets of subject headings will be kept, regardless of the library's profile. When an incoming recon record contains both MeSH and LC subject headings all subject headings on the record will be kept in HOLLIS and display.
For current cataloging a library's locally assigned LC subject headings are tagged as 65X or 69X, depending on the library's HOLLIS profile. If the LC subject headings are to be tagged only as 65X, any 69X on an incoming record is assumed to be incompatible with LCSH and will have its indicator set to "9", meaning that the heading will be keyword-indexed only. If a library is profiled to receive LC-compatible headings as both 65Xs and 69Xs, an incoming heading tagged 69X will have its indicator set to "0", meaning that the heading will be indexed in both the SU and keyword indexes.
For recon records the treatment of subject headings differs according to whether the record is claimed or keyed and is based on the individual library's subject heading practice. Subject headings on claimed records will be indexed as LC subject headings in HOLLIS. Subject headings on keyed records, however will be tagged by the OCLC Retrospective Conversion staff as either local (690 or 691) or LC (650 or 651) depending on the library's cataloging practice. This means that if a library has not consistently used LC subject headings the subject headings on the keyed recon records must be tagged as 69X. When the record comes into HOLLIS the indicator for the subject headings will be set to "9" which will identify them as local headings. The headings will be keyword-indexed only. Subject headings tagged "69X 9" will not undergo OCLC authority control processing.
The sample HOLLIS record below is a keyed recon record containing locally assigned subject headings that are only searchable using keyword.
In the public catalog, these headings are identified by the label "KEYWORD SUBJ." These local subjects (69x, second indicator "9") are indexed only in keyword -- hence the catalog mode label "KEYWORD SUBJ." In fact, all subject headings (6xx) can be searched by keyword, but access to local headings is restricted to keyword; LC and MeSH subject headings are also accessible via the string indexes SU and ME. It is important that librarians inform HOLLIS users that catalog subject searches should include both string and keyword for the most complete subject access.
Within the scope of the retrospective conversion project it is not possible to handle subject headings on a case by case basis and determine whether a subject heading is or is not an LC heading. It will be possible, using the reporting function and global maintenance capabilities, to retag where appropriate certain headings that are known to be LC but have been tagged as local and therefore are indexed only as keyword subjects. Units interested in receiving this type of report should contact the Recon Project Manager.