January 8, 1997
Limiting a Search by Format: New Capabilities
Secondary format characteristics
Technically (for those of you who care), to determine the entire range of format characteristics which apply to a title, HOLLIS is looking at the Format code, LGD fixed field, ARCV fixed field, 006 first position (i.e. "006:") . Since it was not possible to specify more than one kind of format before March 1996, only prospectively coded records will contain this information. The majority of cataloging records for multi-format works will still be retrieval only under their primary format.
Some single values result in more than one format flag being set. The best example occurs in the area of manuscripts.
The qualifier //fo=mss will retrieve materials in the obsolete AMC format, the new Mixed Materials format, materials which have the ARCV fixed field coded as a', and materials which have a Type of Record (LGD position 1) code for manuscripts. For example, a manuscript score, coded "dm" in the LGD, would be retrieved both by an //fo=sco and by an //fo=mss search.
Restricting to primary format
These changes took place in HOLLIS string indexes over Thanksgiving, but they will not be implemented in keyword until the next regularly scheduled regeneration of the keyword indexes, which for the HU database will be Friday, December 13th. This is not a n omen!
Broader Term References
150: 0: $a Aardvark 550/1: 0: $w g $a Tubulidentatapreviously created the index entries:
AARDVARK 1 [authorized heading] TUBULIDENTATA 1 [retrieves related heading: AARDVARK]Now, the same authority fields create 3 index entries:
AARDVARK 1 [authorized heading] 2 [retrieves broader heading: TUBULIDENTATA] TUBULIDENTATA 1 [retrieves related heading: AARDVARK]Unfortunately, HOLLIS cannot insure that the broader term has been used. For this reason, some of the broader term references will only retrieve the reference back to the original term or to the original plus references to other narrower terms, e.g.
TUBULIDENTATA 1 [retrieves related heading: AARDVARK]"Do Not Make" References
HOLLIS has never been sensitive to the coding in authority records which specifies that a reference should not be displayed.
This frequently occurs when the pre-AACR2 form and the valid heading differ only in punctuation or capitalization, that is, when the headings normalize to the same form. In cases like that, the earlier form can be useful in making global changes, but inde x entries built from the reference would be useless.
151: 0: $a Peoria (Ill.) 451/1: 0: $w nnaa $a Peoria, Ill.used to create
PEORIA ILL 1 [authorized heading] 2 [retrieves: PEORIA ILL]Now, the system will not make references if the 4th position of the $w is coded a' - "reference not displayed", so you only see:
PEORIA ILL 1 [authorized heading]CL Searching in BF
Library of Congress classification numbers in the 050 fields in the BF database will be searchable using the FIND CL command. Only LC-assigned 050s will be indexed, and if the 050 field contains more than one classification number, only the first will be indexed, for example,
050/1:00: $a Z1037.8.A7 $b M35 1990 $a PJ7585Results in the index entry
Z 1037.8 A 7 M 35 1990 / 40 malkawi ahmad /tufulah fi isdarat muassasat wa dawr al nashr al/ 1990 bksGPO Item Numbers in the GN Index
Government Printing Office item numbers will be added to the GN index in the HU database. Item numbers are kept in MARC field 074, and have the format 0000, or 0000-A, or 0000-A-00, e.g. 0766-C-03. Other numbers in the government number index now come fro m the 086 (Government Document Call Number, including SuDocs) and the 035:6_:, which is where GPO Shipping List numbers are carried in the shipping list check-in records sent by Bernan. Those numbers have a different structure, and cannot readily be mista ken for GPO item numbers.
HU Numbers in RV Records
FIND LN (MH)AAA1234HU or FIND LN MH_AAA1234HU or FIND LN AAA1234HUThe RV LN index is an exact number index, and cannot be browsed in any meaningful way.
Public Notes will Display to the Public!
151: 0: $a Soviet Union 667/1: : $a The Commonwealth of Independent States is not the jurisdictional successor to the Soviet Union. 680/1: : $i SUBJECT USAGE: For works about this area for the post 1991- period, see the subject headings $a Former Soviet republics; Com-monwealth of Independent States, $i and the names of the individual republics, e.g.$a Russia (Federation). $i For works by or about this area for the 1923-1991 period, the heading $a Soviet Union $i is still valid. $5 WIDwill generate the following index entries in technical services:
SOVIET UNION 1 [authorized heading] 2 [retrieves information on the use of the above heading] 3 [retrieves information on the use of the above heading]while in the public catalog, only this entry will display:
SOVIET UNION 1 [retrieves information on the use of the above heading]Awards Note Searchable In KW
Awards and prizes won by the cataloged title can be recorded in the 586 field. This field now displays in the long display format with the literal "Awards:" or no literal at all, depending on the value in its first indicator. The Awards note will now be s earchable using the KW index. This enhancement will not take effect until December 13th. Known bugs -- soon to be fixed!
There are two known bugs which turned up during these indexing changes:
HOLLIS II Happenings
HOLLIS rate schedule
There will be no changes to HOLLIS fees for next year. (See attached rate sheet.) The expected increase in fee income is based on the continuing trend of increased fee-related activity in HOLLIS, primarily circulation activity. There has also been a dec ision to not introduce any new HOLLIS fees related to the Distributed Reporting system. Rates for reports produced by OIS will remain unchanged.
It is important to keep in mind that we expect that HOLLIS II will be implemented some time during FY99, so based on the current project plan, FY98 will be the final full year of HOLLIS. One issue which you may want to consider for FY98 concerns upgrad es to desktop computers. OIS is collecting data on specifications for desktop machines. We are surveying all potential system vendors in the hopes that we can create a draft profile for desktop computers for the initial implementation of HOLLIS II. Th is may be used to begin planning what desktop system upgrades may be necessary for your library. We hope to have more information available later in December.
Last but not least is the problem of budgetting for reference databases (e.g. AI, LR, BIP, RLIN, etc.) Budgetting and tracking expenses related to these databases have become very complex in recent years. Many libraries maintain their own arrangments wi th vendors, some cooperative deals have been established, and OIS continues to maintain some databases on the mainframe. It is difficult to generalize about these arrangements in terms of their impact on the FY98 budget. If you have specific questions a bout any particular database please feel free to contact Tracey Robinson in OIS.
McGill v.314 available for testing
McGill TCP3270 for Windows version 3.14 is now available from the OIS ftp site at elmer.harvard.edu. Specific steps for downloading are provided below. Note: If you have experienced problems correctly displaying subfield delimiters in HULPR with TCP3270 version 3.1X, this can be corrected by deleting previously installed ALA fonts from your Windows System directory prior to installing Version 3.14. Instr uctions appear below.
The OIS ftp site is not an anonymous ftp site - you will be prompted for a login and password upon connecting. If you do not know the guest ftp login or password feel free to contact OIS.
Directions for Downloading McGill TCP3270 for Windows Ver. 3.14 :
Try this to fix the subfield delimiters...
HOLLIS Circulation Manual new edition released
HOLLIS Newsletter back issues online
HOLLIS Products Winter Delivery
Therefore, it is important to check for the morning messages regarding the HOLLIS products. At least two messages are issued each morning, one for the circulation reports, and one for all the other products such as vouchers and purchase orders. The messa ges tell you when the products are ready, or if some have been delayed. As always, these messages are posted to the common bulletin board of EMS, within HULPR. A quick glance at HULPR can save you time, and supply all the information you need about the HO LLIS batch products. Contact Lauren Caulton in OIS if you have questions.
HOLLIS Plus Gopher
Most of the libraries that will be affected by this switch are using what we call the "common client" for HOLLIS Plus. If you access HOLLIS Plus from a dumb terminal, or by telnetting to hplus.harvard.edu from a registered library devic e and logging in as hplus-xx (where xx is a two-letter departmental code), you are using the common client. If you are one of these users, you will see some small changes. Your login procedures will remain identical. However, because the common client wil l be using lynx (a textual Web browser) rather than gopher, there will be some slight differences in the look and feel of the system once you are logged in--the most obvious of which will be that HOLLIS Plus will suddenly seem to have added quite a number of resources.
**THIS CHANGE WILL HAPPEN ON FRIDAY, DECEMBER 20.**
If this scenario affects you, please take the time to have a look at the new common client, now in place on our test system. You can reach it by telnetting to sirin.harvard.edu and using your regular hplus-xx login ID.
There is a second group of users that will be affected by the demise of the gopher, although there are very few of these users within the library community itself. A few locations access HOLLIS Plus directly via gopher, typically using a desktop gopher application (like Gopher+, WGopher, WSGopher, or TurboGopher). If you are one of these users, you will need to switch to a Web browser for your HOLLIS Plus access, most probably by telnetting to the above-described "common client". If you think this applies to you, and you haven't already spoken with Lydia Ievins about it, please contact her soon for further information. THE GOPHER SERVER FOR HOLLIS PLUS WILL BE REMOVED ON JANUARY 2. We have been working with the library public services departments in order to make this transition as seamless as possible. Please contact Lydia Ievins in OIS if you require any further assistance during this process.
OIS has now obtained, and we will receive monthly, a report of terminals which shows which ones have been used vs. unused each month. The report shows a year's history for each one. We are using this report to disconnect at the UIS end terminals which hav e not been used for more than 6 months. This will allow UIS to move off the machine room floor extra, unnecessary equipment and to clear unused cables out of the way.
If anyone has questions or concerns about this, please contact Linda Marean in OIS.
In recent months, increasing numbers of staff have been calling OIS for information when they have problems accessing the HOLLIS or HULPR systems.
When access to HOLLIS systems is disrupted, a staff member’s first point of contact should be the local network or microcomputer support office. Many apparent HOLLIS problems turn out to be local network outages instead.
If local network support staff cannot help, another option is to contact the UIS Machine Room (495-3000) — if there is a HOLLIS problem, they will know. If the problem does not appear to be local and the Machine Room reports that HOLLIS is healthy, try co ntacting the UIS Help Desk (formerly known as the NIC; 6-2001). Help Desk staff may know of a campus network outage.
If HOLLIS online systems are malfunctioning, OIS will post an announcement to the HULINFO discussion list. Whenever possible, system status and an estimate on availability will be provided. When possible, staff should rely on this notification rather than calling OIS directly. A flood of incoming calls to OIS during systems outages make it harder for systems staff trying to troubleshoot the problem.
There undoubtedly will be times when the only option (during a system outage) is to contact OIS. In that event, we have always recommended that the designated HOLLIS liaison for each unit be the point of contact. That is, staff should report problems to t heir liaison, who should then contact local network support, UIS or OIS.
30,482 Payments approved between 7/01/89 and 12/31/95 have been flagged as weedable. Fiche for these records were distributed last week to those with subscriptions. These fiche should replace the previous fiche (for 7/01/89 to 6/30/95). Any questions can be directed to Linda Marean in OIS.
According to recent feedback, the newly instituted Intra-Library Loan Pilot Project has been very popular with users. On Monday, November 17, there will be an open forum in the Lamont Forum Room to discuss the ILL, exchange ideas, answer questions and obt ain suggestions about this new service. Contact Lee Anne George (496-3096) if you have questions.
In November, the Business School switched to Ovid as the supplier of ABI-Inform. Access to this database via HOLLIS Plus has been disrupted while OIS works with OVID to implement technical changes to enable a seamless log-on process. [In mi d-November it was determined that OVID’s technical changes would be further delayed. OIS implemented a work-around to allow access to ABI-Inform until a proper log-on procedure could be put in place. The database is now available, although the connectio n process is awkward. — Ed.]
Harvard now hosts the online version of the Oxford English Dictionary (second edition; known as OED2). The search engine behind the OED2 is called OpenText; the web interface was designed by the University of Michigan. MacKenzie Smith noted tha t there are no plans to modify the interface this fall. OIS plans to survey users about the interface in early 1997. Library staff are encouraged to use the OED2 and store up comments until the formal review process starts. If you have questions/commen ts that can not wait, contact MacKenzie Smith in OIS.
Chadwyck-Healey’s Literature Online (LiOn) site will be offered on a trial basis from mid-November through the end of December 1996. LiOn offers a variety of full text databases of English and American literature. Staff are encouraged to try LiOn and se nd their comments via the Suggestion Box link from HOLLIS Plus. OIS needs to determine whether Chadwyck-Healey’s interface would be a satisfactory alternative to Harvard developing its own interface.
The demise of the HOLLIS Plus gopher is still on schedule for end of December 1996. See a related story in this issue for more information.
The OIS subnet continues to suffer from a “time out” problem that is affecting web resources managed on OIS servers, including HOLLIS Plus and all OIS-sponsored web sites. The original diagnosis of a faulty piece of hardware in late October prov ed incorrect — response time improved temporarily but began to degrade again in November. OIS will continue to investigate. [In late November, OIS made further hardware changes and is optimistic that this problem has been resolved. If you have ques tions, contact Marty Sasaki in OIS — Ed.]
Tracey Robinson outlined the phases of HOLLIS II project. Phase I (preparation and planning stage; Aug. 1996 - March 1997). Four task groups create lists of critical and highly desirable functions that we want to see in the next generation libra ry system. A HOLLIS II Steering Committee identifies other factors (vendor viability, marketplace status) that should be factors in the evaluation stage. Nine prospective HOLLIS II vendors visit Harvard to demonstrate their systems and learn more about the “Harvard experience.” Near the end of this phase, a short list of 3-4 vendors will be selected for evaluation in the second phase.
Phase II (vendor and system evaluations; March - June 1997). Vendors on the “short list” are thoroughly evaluated by the Steering Committee and Task Groups. There will be return visits by selected vendors and possibly visits by to vendors and i nstalled sites. A system recommendation is expected in June 1997 and formal review and University approval are planned for the July-October 1997 timeframe.
Phase III (implementation; June 1997 - Jan. 1999). Planning for the transition starts at the end of Phase II (June 1996); implementation begins in early 1998 and lasts through early 1999.
Is there enough information? Tracey Robinson asked whether the amount of information regarding HOLLIS II activities was satisfactory. The general response was that OIS should continue providing as much information as possible and that too much informati on was better than too little.
There is a HOLLIS II web site:
with up-to-date information on planning activities. Periodic progress reports also appear on the HULINFO electronic mail discussion list. In the future OIS expects to schedule more open meetings to provide progress reports. There will also be occasiona l inserts in the print publication Harvard University Library Notes. Anyone with questions or suggestions regarding spreading the word on HOLLIS II should contact Julie Wetherill in OIS.
According to Recon Project Manager Karen Carlson Young, a whopping 4,325,000 records have been transferred to HOLLIS since the Recon project’s inception. Incredibly, the last of the Widener and Houghton records (approximately 2.2 million) were processed i n early November. Projects still in progress include the addition of Tozzer records to the AL database, serials from Divinity and Widener. Processing of Widener’s non-Roman holdings (Slavic and Middle-Eastern) is being done in-house. A small CJK project t o convert Law School holdings is being done by RetroLink. Harvard-Yenching’s CJK records are being converted by OCLC in a separate project that is expected to last another five years.
Karen announced that the end of the entire RECON project is anticipated for December 31, 1996. In addition, the OW database will be removed from the public catalog by the end of the current semester. [OW is likely to remain in tech. services mode thr oughout 1997 for ongoing cleanup--Ed.] Financial reports and heading corrections will continue to be produced. As part of Harvard’s Recon contract, OCLC has been providing authority control for Recon and current cataloging. Since this contract is va lid through October of 1998, OCLC will continue to provide this service to Harvard through the end of the contract. As yet, there are no plans for beyond that point. Contact Robin Wendler if you have question s.
Robin Wendler reviewed for liaisons changes to the HOLLIS enhancements policy that were recently approved by HAAC. A full description of this new policy appeared in the October HOLLIS Newsletter and will not be reproduced here.
rface to HOLLIS. Tracey replied that OIS expects to support these devices through the end of the current HOLLIS system (sometime in the 1999 time frame). Whether these devices can transition to HOLLIS II is no yet clear.
Robin Wendler reviewed the latest developments in format integration. A full description of these features begins at the top of this document.
ABCD WWW Working Group mtg.
3 January 1997
HOLLIS II Vendor open mtg: Ex Libris
Technical Services Managers Meeting