Library Technology Guides

---

Volume 2 Number 06 (June 1982)

CLSI to discontinue 11/04 software enhancement

A number of libraries with CLSI LIBS 100 systems have received letters over the past few months advising them that the company will discontinue software enhancement for systems with PDP 11/04 processors after release 26.

This means that while the libraries will continue to receive hardware and software maintenance, they will not get enhancements or improvements in software after mid-1983 when release 27 is issued.

The effect of the CLSI decision is to freeze the development of all of the PUP 11/04 configured systems and to apply pressure on the libraries to "swap-out" their processors for the larger, more modern PDP 11/34. However, libraries which consider this option may wish to wait until CLSI offers the option of a PDP 11/44 configuration. The PDP 11/44 is a larger, faster and more recently introduced machine than the 11/34 and costs just $6,000 more.

This is, as far as the editors recall, the first time that a turnkey vendor has announced a discontinuation of software enhancement.

Desk-top microform cameras unveiled

The National Micrographics Association Conference is always a major event. Most companies introduce their new micro-graphics equipment at this time. The 1982 Conference held in St. Louis recently saw the introduction of several new microform cameras, all of the type that Fuji likes to call the "No-paper Copier."

All of the machines are approximately the same size and cost approximately the same as office paper copiers. They are also equally simple to operate. The output usually consists of strips of microfilm that can be inserted into microfiche jackets. A sixty-frame jacket costs approximately $.50 to produce. The cost of the supplies themselves is less than a half cent per original page. Most of the machines can copy 1200 to 2500 pages an hour. The maximum page size is 8-1/2 by 14 inches. The machines require no special wiring, plumbing, or lighting. [Contact: Fuji, 350 Fifth Ave., NY, NY 10182; Phillips Company, W 255 N499 Grandview Blvd., Waukesha, WI 53186; 3M Co., P.O. Box 33600, St. Paul, MN 55133; Alos Micrographics, P.O. Box 407, Walden, NY 12586.]

COM continues to be viable

NMA Conference attendees participated in a number of sessions on COM (computer output microform). The speakers generally forecast continued viability for the medium throughout the eighties.

Other forms of storage such as optical disk, magnetic memory in chip or board form and magnetic disk can provide faster access than COM, but they are also far more expensive per megabyte of storage.

Probably the most critical area to examine when comparing COM to other technologies is that of data base size. CON is most appropriate for small data bases say the experts. A large number of CON applications fall into the range of four to six megabytes of data or less, and 90 percent of all COM data bases contain less than 200 megabytes (approximately 100 microfiches).

Handling ease is a factor which mitigates against the creation of very large fiche files. Fiche Tiles become difficult to use when there are several hundred fiches. Microfilm is not the answer because hardware costs rise dramatically when a single file requires more than one microfilm reader.

COM is a good choice when cost is an important consideration. The cost per megabyte of COM has been estimated to be 65 cents as compared with $20 for magnetic tape storage. It would be prohibitively expensive to put a 30 or 40 megabyte database on an optical disk, or some alternative emerging electronic medium, due to the cost of creation, duplication and storage. By 1983 the cost per megabyte for optical disk storage is expected to be $80 but this picture may change dramatically by the mid 1980s.

CAR use grows

CAR (computer assisted retrieval) of information stored on microform is the fastest growing segment of the micrographics market according to the speakers at the annual meeting of the National Micrographics Association.

The micrographics medium is still the most cost-effective method of filing or storing information and there is little reason to believe that this will change during the 1980s. The use of a computer to access and retrieve information stored on microform has combined the low cost of microform storage with the rapid retrieval that is deemed essential in today's offices and libraries.

CAR systems exist for both roll microfilm and for microfiche. A typical 16mm roll film CAR system might consist of:

• An intelligent microfilm reader-printer.
• A mainframe or minicomputer with one or more CRTs linked to the reader-printer.
• Cartridges of film containing document images.

The intelligent reader-printer becomes a computer peripheral for document search and display; the cartridges of film the secondary data base.

In a typical CAR system, relevant summary data is extracted from each document after it has been filmed. A coded address for each document is added to this data base at the time of data entry. Tn a search for information, the user has two choices:

1. Access the computer's primary data base via a CRT for summary information-satisfying a majority of the inquiries;
2. Access the secondary microfilm data base via the intelligent reader-printer for the entire record satisfying the remainder of the inquiries.

Each document in a CAR system is assigned a physical storage location, or address, by a reference number related to its position on a roll of film; or in the case of microfiche, a grid location. Since documents are normally added to the system in batches and randomly stored, the input sequence does not necessarily resemble the sequence in which they are used. Additionally, the search parameters available to the user are not necessarily related to the physical document storage location.

The power of a CAR system is in the merging of computer and micrographics technologies, providing the ability to identify a document from its descriptors, locate it in a secondary data base of thousands of documents, and display it in seconds.

With the appropriate search software, CAR systems can also be designed to handle complex inquiries that ask the system to display a document or documents from multiple search parameters.

Almost all of the major microform equipment companies offer CAR systems. The best sources for articles and ads are Information and Records Management and Information Manager.

Digital telefacsimile in libraries: Denver Public and others?

The new generation of telefacsimile equipment-digital machines which permit the transmission of a page of copy in less than a minute-has been the subject of two items in previous issues of LSN (Vol 1, No 2, pp. 10-11; and Vol 1, No 4, p 40). Viewed in library applications primarily as a rapid and economical method of document delivery in middle range to long distance interlibrary loan situations, digital facsimile is being harnessed in an innovative local application in the Denver Public Library

Concerned to maintain the level of reference service in branch libraries in the face of depleted staff resources, Denver Public has installed Rapicom 6300 digital telefacsimile equipment in the central library and Rapicom SOC's in seventeen of its branch libraries. The machines transmit a page of information in from 15 to 45 seconds. The network enables staff at the central site to rapidly deliver to the branches materials which meet reference inquiries placed at the branch libraries.

In providing the rapid delivery service, Denver Public supplies the first three pages of copy gratis and levies a fifty cent per page charge on subsequent pages of information. Since the system began operation earlier this year, a number of users have opted to pay the copy charge in preference to traveling downtown to consult the material free of charge.

That telefacsimile is undergoing a renaissance is evident from the sales figures-while there were only 255,000 machines installed in North America in 1980, 48,000 units were delivered in 1981. This growth can be expected to continue as the fifteen or so companies now in the market expand their product lines and are joined by new competitors.

The spread of this technology to libraries may be hampered by memories of the limited success of slower (4-6 minutes per page), older generation analog equipment of the 1960s. In assessing the appropriateness of digital telefacsimile in a particular library/library network situation considerations should include:

• establishing a time standard of no more than 60 seconds for the transmission of normal copy. A machine which operates at a slower speed will result in unacceptably high telecommunications charges.
• specifying digital telefacsimile equipment. Not all subminute telefacsimile equipment is digital. However, digital machines combine the advantages of rapid transmission with superior resolution capabilities. While such capabilities are not critical in the transmission of regular print pages, they are essential when complex charts and copies of older published material are being sent. The typeset letters in older materials are often quite close together and this, together with the discoloration and shrinkage in paper than occurs over time, can make an image transmitted on lower resolution analog equipment unreadable.
• specifying equipment which meets the Consultative Committee on International Telephone and Telegraph (CCITT) Group III standard. Adherence to this standard will usually ensure compatibility among the digital machines of different manufacturers. If a library wishes to ensure that its digital machine is also compatible with analog equipment, it should also specify downward compatibility to Group I and/or II machines.
• investigating the possibility of leasing/renting the equipment rather than purchasing it outright. The telefacsimile market is undergoing rapid change. Even though most of the digital machines have been introduced within the past three years, a fourth generation of equipment is under development. The new machines are expected to be capable of faster transmission with greater resolution than now available. The new generation machines will call for satellite, microwave or coaxial cable transmission. Standard development for such machines has not yet been completed.

There is also evidence that a number of photocopier manufacturers are experimenting with the introduction of copying machines with telefacsimile capability. The machines could be used for conventional copying when not in use for telefacsimile transmission.

Videotex or teletext?

At the recent Viewtext 82 conference in New York some industry analysts predicted that, in the immediate future, costs and the regulatory climate would favor the development of teletext rather than videotex systems in the United States.

Both systems display viewer-selected print and graphics information on modified television receivers. They differ in the way that this information is delivered:

teletext systems piggy-back digital signals on regular TV broadcast channels; videotex is a two-way system which requires land-based cables such as telephone lines or coaxial cables to link the user to the central host computer. Videotex systems permit two-way communications between the user and the system; teletext provides only one-way access- the user can select from the available information but usually cannot input data to the system. When used in conjunction with a touchtone telephone some limited interaction can be achieved on a teletext system.

Predictions of the degree of market penetration of cable television suggest that it will not be sufficient to support extensive videotex systems until the last decade of this century. To date, most of the experimental videotex systems in the United States have utilized telephone connections to link users to a host computer. While the recent Justice Department settlement with AT&T will permit the company to participate fully in videotex services, there is a strong possibility that other companies with information interests might seek to limit this activity as anticompetitive. Recent and continuing rises in telephone charges could well be a further constraint to the growth of videotex systems.

Potential information and service providers are not backing off, however. The recent announcement of Time Inc.'s upcoming teletext experiment may signal a change of focus. Scheduled to begin in the Fall of 1982, Time's test will be undertaken in San Diego, CA, and Orlando, FL. The Time test further details of which will be released this summer-is unique in that it will be the first to devote the capacity of a full television channel to the medium, rather than using the limited vertical blanking interval usually between frames normally employed for teletext services. It is projected that this will provide the capability of han-dling some 4000 pages of information, giving the system the appearance of being more interactive than those with the re-stricted blanking interval capacities of just 100 pages. Officers of Time Inc. say that their test will focus more on content than have previous services and that attention will be paid to the development of advertisements that fit the new tech-nology. Time plans to generate revenue by charging advertisers, permitting the end user to access information free or for a minimal charge.

Not all signs point to a demise of videotex. Keycom Electronic Publishing, a joint venture which includes Honeywell, Inc., plans to offer a hybrid videotex and teletext service in Chicago by mid 1983. The project will utilize standard telephone lines for the videotex portion of the service.

Despite assessments which include statements such as "videotex (has] been an utter failure so far," frequent comments about videotex and teletext being technologies in search of a market, charges that the media are elitist both in terms of required literacy levels and financial resources, and opposing views on the degree to which the absence of standards is limiting or encouraging market penetration, videotex and tele-text are far from dead. Over 50 tele-text and videotex experiments are currently underway. Currently reaching an estimated 40,000 households, television captioning for the hearing impaired is the most widely used teletext service in the country. Other signs of continuing interest within the United States include the introduction of a Boston-based American node for Britain's commercial videotex service-Prestel-and the establishment of the Videotex Industry Association to promote and encourage the development of videotex and teletext in the United States. The Videotex Industry Association may be contacted at 2000 L Street, N.W., Suite 200, Washington, DC 20036. (202) 544-5655.

New telephone numbers for CTI

Computer Translations Inc. (CTI), 1455 South State Street, Orem, Utah 84057, has a new telephone number 801-244-1390 which gives direct access into the Library Systems Division. In addition, CTI has installed an incoming nationwide (excluding Utah) WATTS line. This number is 1-800-453-1195.

Carrollton press cluster plan

Carrollton Press, the vendor of REMARC, has introduced a new pricing plan to promote retrospective conversion by "clusters" of cooperating local and regional libraries. A cluster's first hit on a title in the REMARC data base will be charged at the regular rate of $.50, subsequent hits by libraries in the group will cost $.20 per hit. A group using the "cluster plan" does not require a minimum number of participants nor a minimum volume of transactions.

Carrollton Press also offers clusters cost savings in matching records against MARC tapes for $.15 per hit, compared with the individual library rate of $.20 per hit.

Recognizing that the negotiations to create multitype library clusters might take months or even years, Carrollton will negotiate REMARC contracts with individual libraries and permit these libraries to be absorbed into cluster contracts later. When the cluster is established each library will receive full credit for the first-time-use records it had acquired and future uses of these records by the members of the cluster will be charged at the reduced rate.

As further encouragement for libraries to undertake retrospective conversion using REMARC, Carrollton will loan libraries microcomputers, free of charge, to be used in keying abbreviated shelf list records onto floppy discs. The discs will be sent to Carrollton for matching against the REMARC data base.

[Contact: Carrollton Press, 1911 N. Fort Myer Drive, Arlington, VA 22209 (800) 368-3008.]

Online online

The publisher of Online has gone online. The Online Chronicle mounted on Dialog, covers databases, online services, terminals, printers, micros, corporate news, and the information utilities such as Compuserve and The Source. Much of the material from The Chronicle will not have any print counterpart. The file is priced at $35 per connect-hour, plus $.15 per typed record, or $.30 per offline print.

Patron access catalogs- some considerations

Librarians investigating the selection of automated library systems are increasingly encountering references to PAC's (public access catalogs). PAC's- online catalogs which can be accessed by library users without assistance from staff are still very much in the developmental stage. Even so, some fifty libraries now have some form of patron access catalog, approximately half of which are mounted on minicomputer based turnkey systems.

Although there is not yet an established body of opinion as to what constitutes a patron access catalog (PAC) the following elements-in addition to those which normally apply in the selection of an automated system-deserve consideration:

• Data base design and content

  The data base which supports a patron access catalog need not necessarily mirror the contents and practices of card catalogs which frequently contain only a limited selection of the items held in a library. In a PAC data base not all formats might be represented by equally long records. A library should decide what it wishes to include and ensure that the appropriate formats can be accommodated.

  The system should be able to accept, store and reproduce full-MARC bibliographic records. Since the cost of maintaining full records on-line will be significant and some libraries may not feel that they need them, the system should also be able to accommodate partial records that are MARC compatible.

  To facilitate the input of original records, the system should provide an appropriate MARC work form display and a substantial amount of input support and error checking.The system should permit a library to select the specific fields it wishes to include in its data base and to include an adequate amount of information in each.

• Authority control

  Subject, name and series authority files must be available online and must be capable of being maintained on a current basis. The system should support a fully automated, interactive authority control system in which each heading in the authority files is linked to each occurrence of that heading in the bibliographic file, permitting automatic updating of all occurrences of a heading through a single machine transaction.

• Number of access keys

  The number of access keys-the points through which information can be retrieved-should, at least, be as numerous as the access points available to users of a card or CON (computer-output- microform) catalog. Theoretically the number of keys may be unlimited, but practically there is a point beyond which the cost in terms of system performance and the amount storage for indices becomes unacceptable. A library should determine the limit for its particular system and compare that with a list of the access keys it desires.

  It should not be necessary to use derived or algorithmic keys made up of combinations of letters from author's names and the title or other elements of the entry. Such an approach is unacceptable in an on-line catalog designed for patrons with little experience in using computer terminals.

• Subject access

  A patron access catalog should offer substantially better subject access than the 1.5 average Library of Congress subject headings typically associated with each record in a traditional catalog. Free-text searching of title and subject fields should also be available. The system should permit Boolean searches using AND, OR, and NOT relationships as this technique offers a dramatic improvement in access over card or COM catalogs.

• User cordiality

  One of the major reasons for providing patron access catalogs is to promote the identification and location of information by patrons without the assistance of library staff. To do this successfully the system must be easy to use. This can be achieved by careful selection and development of the hardware and software. Terminal keys can be color coded or touch terminals can be used so that the user does not have to search a keyboard for the right keys. Software can be written to provide prompts to guide the user.

  There are disadvantages to these easy-to-use terminals: the experienced user may become impatient with the deliberate and gradual nature of the prompts. An ideal system would employ terminals with keyboards and would offer bath prompt and command modes to that experienced users could skip the prompts and fully control the system.

• Display and print formats

  A PAC needs a variety of display and print formats. Displays should include a format that resembles the catalog card and other formats which emphasize the portions of a record that are of primary concern to the library's clientele. The system should provide an appropriate method of displaying multiple records when a search results in more than one hit. An abbreviated format will allow the simultaneous display of several records, but it should be extensive enough to facilitate meaningful comparisons. The entire record should be available upon a simple command. The appropriate display for a "full" record for a patron will be different from that for a cataloger. The system should allow the user to browse or page forward or backward to examine related entries.

  The terminals should be able to support printers. Printers normally reproduce that which is displayed on the screen. The use of a printer with a terminal that displays only full bibliographic records would tie up terminals and might force patrons to wait too long. If the software does not permit the printout to be limited to a single line of information per record, printing should be done off-line and only when specifically requested by the patron.

• Number of terminals

  At this stage there is little evidence on which to estimate the number of on-line terminals a library should install for its public access catalog. PACs which offer sophisticated searching capabilities such as Boolean logic will probably require more terminals than the number of COM readers needed to support a microform catalog. One CON reader for each 200 patrons using the library in a day is an accepted rule of thumb for most libraries: an on-line catalog might require one for each 100 daily users. A library might size the system on the 1:100 formula and budget on that basis, but install on the basis of 1:200 and use the system's auditing and reporting features to adjust the number to the correct level.

  This list of elements, supplemented by those which apply to all automated systems-back-up, transferability of software, capacity for expansion, five year cost, etc. should be taken into account when planning or evaluating patron access catalogs. The list will change over the next year as more libraries gain more experience with PACs. The professional literature should be followed carefully and those planning to implement patron access catalogs should seek to correspond with librarians at the institutions which have already launched their efforts.

---

Maintained by Marshall Breeding

Copyright 1994-2024