Library Technology Guides

---

Volume 6 Number 12 (December 1986)

The automated library systems industry: a time of change

Anyone observing the automated library systems industry since CL Systems Inc., the forerunner of CLSI, installed its first system in 1973, has seen a great deal of change. More than one thousand libraries have installed systems successfully, and many more hundreds of libraries will be doing so in the coming years. Given the volatility of the marketplace, some of these latter libraries likely will encounter difficulties as have their counterparts in the past. One kind of information which might be of interest to librarians embarking on either a first-time system purchase or an upgrade is the activity taking place within the industry with regard to acquisitions and major financial investments. The decisions being made in corporate boardrooms today surely will have some effect on both vendors and libraries.

Some three years ago, OCLC acquired Avatar, a small company which was enhancing and marketing software developed at the Lister Hill Center for Biomedical Communication. That product later became OCLC's LS/2000 local library system. OCLC subsequently acquired Metamicro, a micro-based software product which later evolved into OCLCs SC350 serials control system. This seems to have been one of the first efforts within the industry to combine parts of different companies to create a single product line.

A year later, The International Thomson Company (ITO) acquired Carrollton Press along with Carrollton Press' work on the Library of Congress REMARC product, which provided them with access to a very large machine-readable bibliographic data base of materials cataloged by the Library of Congress before its MARC tapes service became available in 1968. The company further enhanced its library position with the acquisition of UTLAS, thereby adding the capacity to access, via both off-line and online methods, a combined data base that is both current and retrospective. ITO then proceeded, through UTLAS, to buy the ALIS III library automation product from Data Phase Systems Inc. [UTLAS bought the rights to the ALIS III software, which it has renamed T50, and has entered into negotiation with former. Data Phase customers. Contrary to a popular perception, UTLAS bought neither Data Phase itself nor-its-customer base.] Through these purchases, ITO is constructing a library business of complementary components, rather than adhering to a more traditional and narrowly defined role of local systems vendor.

Within the past couple of years, TBG, a multinational family of businesses headquartered in Europe and with a U.S. office in New York, purchased BRS, which provided TBG with both an online data base and search software. The latter has become an ingredient in many different systems, including some of the emerging CD-ROM-based systems. TBG subsequently purchased CLSI, the largest local systems company in the library automation industry. TBG also had been negotiating with Northwestern University to acquire the NOTIS product, but the proposed purchase is no longer under active consideration.

Like ITO, TBG seems to be assembling a broader range of products and capabilities than the industry has seen in the past. Both companies, in fact, have acknowledged their intent to position themselves strategically as "information" companies which they perceive as having a sound future. To implement that strategy, the companies appear to have undertaken a building block approach of which the acquisitions made to date presumably are only a small part.

The acquisition of both Dynix and the rights to the Colorado Academic Research Library (CARL) software package by the Eyring Corporation, heretofore a defense contractor with a high-technology engineered software development house, apparently was made with the intent of offering automated library systems at the lower and middle segments of the market , via Dynix and at the upper end of the market via CARL. If, indeed that was Eyring's strategy, the plan seems now to have changed. Recent reports suggested that Eyring would sell Dynix to a major computer company. In fact, in mid-November 1986, the management of Dynix arranged a buyout, thus making Dynix an independent company. It has been reported that, while a major outside investor was involved, the principals in Dynix control the majority of the voting stock.

In addition to the activity already outlined, Gaylord Library Systems, a division of Gaylord Brothers, Inc., of Syracuse, New York, acquired Library Systems and Services, Inc. (LSSI) , thereby providing itself access to the bibliographic service market. Gaylord historically has offered local library systems products, primarily in circulation, using a distributed processing approach and a number of micro-based products. Those who still think of Gaylord in terms of book pockets, mending tape, and the old Gaylord, Circulation Machine, must realize that it is also a multimillion dollar library automation company. There have been indications that the company might be interested in positioning itself further in the library automation marketplace. Several years ago, it was reported that the company had entered into preliminary negotiations to purchase the marketing rights to one of the standalone turnkey systems but nothing ever came of that particular deal. However, one cannot rule out the possibility of another similar acquisition or the development of a new standalone product since money for those purposes seems to be available.

A number of computer companies recently have increased their involvement in library systems. McDonald Douglas has established a library subsidiary to market its Urica product. Sperry has been bidding more aggressively not only the PALS software, but also software developed by SIRSI, a small Huntsville, Alabama-based company which has not previously had the strength to market outside of the Southeast. IBM has reentered the library marketplace after a short absence, with a revitalized DOBIS product. IBM has Americanized the product, and the individuals primarily responsible for software development have moved from Germany to Connecticut. With development work now based in the United States, the DOBIS product likely will be marketed more vigorously.

Major capital investments beyond acquisitions have also been made recently within the library automation industry. INLEX, a California-based turnkey vendor that markets a system which operates on Hewlett-Packard equipment, has become much more visible due to an infusion of money which has made it possible to expand both its software development and marketing staffs. The Carlyle Corporation recently has benefited from an infusion of approximately two million dollars in venture capital which has allowed it both to pay its debts and to invest in development and marketing staffs.

On the downside, Data Phase was forced to scale back its operation and BLIS has ceased operations. The Geac Corporation after going public on the Toronto Stock Exchange in order to raise capital, has since that time, faced serious difficulties stemming mainly from the dramatic changes in the banking industry. The result has been a significant drop in its stock price, a period of time when trading of the stock was halted, a restructuring of its debt to satisfy lenders and creditors and a corporate management shake-up. It is too soon to suggest what long term impact this will have on GEAC both as a corporation and as a library systems vendor, but the short term prospects appear to be trouble-some, at the least.

When considering the number of acquisitions that have taken place in a relatively short period of time, it appears that the major implications are financial. In virtually every case, the changes have resulted in new capital for organizations which had been constrained by limited financial resources. The infusion of capital has provided money to strengthen both development and marketing efforts.

A perhaps unappreciated effect of the increased capitalization has been to strengthen vendors' ability to secure performance bonds. As a library seeks a performance bond, either by choice or as mandated by the community or organization of which it is a part, a vendor complies by obtaining a bond through a bonding agent which assesses a fee based on a percentage of the value of a project and an assessment of the risk involved. The cost is passed along to the library. The less secure a company is in the eyes of the bonding agent; the higher the price of the bond, which in turn raises the price to the customer even further. In some cases, a bonding agent will refuse to issue a bond at any price. To a bonding agent, the infusion of capital ostensibly enhances a vendor's ability to perform which enables it to secure a bond at a lower price, thus permitting it to reach a greater part of the market which otherwise would be unwilling or unable to purchase the system due to the vendor's higher price.

Capital also enables a vendor to offer a library a wider range of financing options. Normally, vendors sell a system.

If a library were to lease a system, it ultimately does it through a third party leasing company. After having sold the system, the vendor sells the "paper" to the leasing company which pays the contract amount less a specified percentage. The library then has a contractual relationship with the leasing company. As a practical matter, this makes the kinds of remedial holdbacks and other techniques used as leverage to ensure vendor performance difficult to implement.

A company with sufficient capital can undertake a variety of financing approaches using its own resources rather than those of a third party. This enables it to realize the full undiscounted value of a sale and to offer greater flexibility in terms of customer payment schedules. An innovative example of this is employed by CLSI. Through CLSI's "unit pricing" concept, a library can contract for a phased system which ultimately will include a prescribed number of terminals, but which initially would include fewer than the ultimate number. At the initial stages of system implementation, the library's purchase is financed by CLSI, though after a break-even number of terminals has been installed, the revenue from the project begins to contribute to the company's profit. In order for this to succeed, there must be an established price and a date at which the library is obligated to bring the system to its prescribed size. Until that date, however, the library pays only for the number of terminals actually installed.

Aside from the financial considerations discussed above, other issues may arise for companies which become part of new organizations or have new investors. Among these are financial, management and. legal reviews which often result in a restructuring of the company financially, a change in the management structure and personnel, a change in the way decisions are made, a change in the emphasis and direction of the company away from product development and towards marketing efforts, etc. Each of these could be analyzed in detail, but more important at this point is an understanding of the direct impact that these changes could have on libraries.

In the long run, the changes taking place in the library automation industry are apt to result in the emergence of fewer but more financially viable companies. As companies stabilize and establish market niches, it becomes more difficult to find investors willing to challenge the status quo either through an acquisition or through a new venture. As a result, the prospects for further changes decline. While this might enhance a library's confidence level in terms of an individual vendor's ability to remain viable during the life of a library's system, it also might result in a reduction in the kind and pace of product development, given the reduction in the marketplace competition. Conversely, one might surmise that once financial constraints are removed, libraries could expect a more predictable schedule of software development and enhancement. Each conclusion has its proponents, yet only time will tell what will be the final outcome, though one certain outcome will be continued change.

Another aspect to this cycle is the increasing likelihood of library system becoming more like commodity products. Sooner or later, the level of maturity which has been achieved in the circulation product apparently will be achieved in acquisitions, serials control, and patron access catalog modules. When this happens, companies no longer will sell against one another on the basis of functionality, but instead will compete on price, postsales support and other non-product-specific elements. Products will move toward maturity more quickly given the consolidation of and investment in the industry, and libraries often will have to evaluate systems based on a range of elements beyond functionality.

Another change libraries reasonably can expect is an increase in software prices. One outcome of the types of management and financial reviews mentioned earlier is an increased emphasis on the value-added nature of the software. In today's marketplace, vendors are realizing greater margins on the sale of manufacturers' hardware than on their own proprietary software. As the market changes, so too will that emphasis. Hardware margins already are falling and vendors are beginning to price software nearer to its value. This does not mean necessarily that the total price of systems will rise, as one might see a commensurate lowering of hardware prices along with higher software prices, yet in the course of this transition an overall increase should be expected.

Finally, these changes are apt to result in more difficult contract negotiations as vendors become less willing to deviate from their standard contracts. The hardening of positions already has begun. The legal reviews underway in companies are encouraging vendors to make contractual terms and conditions measurable. Ambiguities and special contractual provisions increasingly will fall by the wayside, as vendors become more sensitive to the wording of their contracts. In their desire to maintain their standard contract, vendors will be less willing to agree to changes during negotiations without first consulting attorneys. Vendors also may seek to introduce "best effort" clauses, rather than unconditionally committing to functionality or performance.

Contract negotiations will remain the most critical aspect of the system procurement process. That is not to suggest that a contract negotiation need be an adversarial process. Rather, in the best of circumstances, one should prefer an agreement which can be filed away and forgotten after signing. However, the contract is a contingency document to be used if the face-to-face relationship with a vendor breaks down. And, at that point, the contract had better be a document which equitably and fairly protects both sides.

In any case, for good or for bad, library automation has already moved into the realm of big business.

OCLC to create cooperative data bases

OCLC has announced the availability of a new program through which it will create and maintain data bases of bibliographic records for various groups of institutions. Bibliographic records from the so-called state data base project would be resident in the OCLC Online Union Catalog (OLUC). These data bases would be integrated into the OLUC and accessible by all OCLC users. In addition, a profiling process is available to identify state data base participants and link, their bibliographic records and holdings. This would provide state participants with access both to information defined on a geographic basis and to the national OLUC for needs which are not met within the state.

Though the name implies that the common denominator for participation in this project is all libraries within a state, in fact, any group of libraries may band together to participate. For example, consortia whose members either cross state lines or are within a single state though they comprise fewer than all libraries in the entire state, are eligible for participation in such an OCLC project. For example, a group of California public libraries currently is developing a state data base and in North Carolina a comprehensive data base, whose participants include all types and sizes of libraries, is being created.

Records for creation of state data bases can be contributed in two ways, via online system or tape loading activity. Online system activity, including historical and ongoing cataloging and retrospective conversion, adds records to the state data base. A profiling process links all records into a single state data base. Much of the data base, therefore, is built automatically as a result of the online OCLC-based cataloging and retrospective conversion performed by the libraries. Machine-readable records from other sources also can be batch loaded into the OLUC and the state data base via tape loading.

OCLC members may access the system via all standard OCLC record access points, during the regular OCLC hours of availability. Tapeloading libraries may access OCLC using all capabilities extended to current members and may perform searches of state data base holdings.

Some libraries may not find it practical to contribute records to the state data base, either through online activity or through tapeloading. These libraries, however, may require access to the state data base for resource sharing or other purposes, and the libraries, called "selective users" (in that they have selected a single, non-cataloging use of OCLC service to use), may freely search a state data base. Selective users access abbreviated bibliographic records from the entire OCLC and state data base holdings. Selective users participate in state data base activities, but do not have access to holdings outside of the state, or group, in the case of a multi-state consortium.

OCLC supports resource sharing for all state data base participants and holdings information for the data base is available to all data base participants. Use of the Interlibrary Loan Subsystem is available to OCLC members, tapeloading libraries, selective users and referral centers.

[Contact: OCLC, Inc., 6565 Frantz Road, Dublin, OH 43017-0702; (614) 764-6000.]

Question from a reader

From time to time, the editors receive questions from LSN readers. When questions do arise, we make every effort to respond. A question arrived recently from a reader in Colorado who asked, "Do you have knowledge of any UNIX-based library automation software for any micro, supermicro, or minicomputer? Do you know of any libraries or developers that use Sun supermicros?"

There is a UNIX-based automated library system, called Unicorn, sold by SIRSI, a Huntsville, Alabama-based, privately held company founded in the fall of 1979 for the purpose of developing and selling library automation products. SIRSI's 1985 sales were approximately $500,000, and they realized a small after tax profit. The company claims that as of August 1986 sales were already in excess of a half million dollars.

The Unicorn Collection Management System is a turnkey automated library system originally developed by SIRSI for the Georgia Institute of Technology. Unicorn uses the UNIX operating system and thus is able to run on any UNIX host computer, intelligent terminal or microcomputer used as a workstation. It is written in the C application language.

SIRSI designed Unicorn as a distributed intelligence system to enhance efficiency and response time, while reducing hardware storage requirements and cost. The bibliographic, item, and patron data bases are stored on a host computer, where all Unicorn transactions are processed. The host computer is only used when it is necessary to communicate with the data base. Operations such as selecting a command, editing a screen, wanding labels or entering information are performed entirely on an intelligent terminal (often referred to as a workstation in SIRSI product literature) or microcomputer-based workstation. This saves host computer time by transferring some of the load to the remote peripherals.

SIRSI initially developed Unicorn in 1981 and began general marketing in 1983. By mid-1986, the company was supporting 13 completed installations serving a total of 24 libraries. Included in the customer base are two regional library systems each consisting of three libraries and another system shared by eight libraries. The other ten installations each serve a single library site. With the exception of a single site in Arizona, all SIRSI systems are installed in the southeastern United States, including North Carolina (2) , Georgia (6) , Florida (1) Virginia (2) and Tennessee (1). The customer base includes a mix of public, academic and special libraries, but virtually all the libraries are small. The sales for which information is available include 13 systems of fewer than nine terminals, three systems of between nine and 16 terminals and one system with more than 17 terminals.

As for the second question regarding the use of Sun equipment, the editors are not aware of any such library installations or system developers using Sun supermicros.

[Contact: SIRSI Corporation, 8106-B South Memorial Parkway, Huntsville, AL 35802; (205) 881-2140]

OCLC profiles its LS/2000 user community

OCLC now has a total of 76 staff working in the LS/2000 division of which 17 percent work in administrative/financial/secretarial functions, 12 percent in marketing, 35 percent in implementation (supporting activities between contract signing and acceptance of a system) 24 percent in support (hardware and software support and the user support center to answer telephone inquiries from users), and 12 percent in development. Spies noted that the number in the development function need not be as large as at some other vendors because they rely on other OCLC staff as well as out-of-house development people for work in this area.

Of the 87 libraries who use LS/2000, 8 percent are public, 10 percent medical, 26 percent government and 5 percent other. OCLC's contracts in the past two years, especially the University of Wisconsin (11 systems serving 11 campuses) and the University of North Carolina (9 systems serving 12 campuses) have added significantly to the number of colleges and universities choosing the LS/2000 system. Academic libraries now constitute 51 percent of the LS/2000 community.

OCLC reports that the libraries who purchased the LS/2000 system in the past year are quite similar in size (both in number of terminals and bibliographic records) to those libraries who purchased the LS/2000 system in previous years. In fall 1985, 54 percent of the LS/2000 sites had 16 or fewer terminals. Today, 58 percent of the 15/2000 sites have 16 or fewer terminals. It is important to note that these numbers in some cases reflect the initial installation of terminals and not the ultimate number of terminals on a system. In his 1985 automated library systems article (Library Journal, April 1, 1986) Joe Matthews reported that at the end of 1985, 67 percent of all automated library systems had 16 or fewer terminals. OCLC's 58 percent would indicate that a greater number of LS/2000 sites are larger than the norm. In fall 1985, 56 percent of the LS/2000 sites had data bases under 100,000 records. In fall 1986, that figure remained the same. Fewer than twenty users have a data base under 20,000 bibliographic records, between 30 and 35 installations each have between 20,000 and 100,000 and between 100,000 and 500,000 records respectively, and fewer than five have greater than 500,000 records.

In another area of interest to LS/2000 users, OCLC reports that the following new features are being added to the system's authorities capability: the ability to match an internal LS/2000 authority record generated from a bibliographic load against an external MARC authority file; the ability to add information from the external file to these records; the ability to create additional cross-references and add note information to the internal authority record; and the ability to display subheadings and notes from these records in the online catalog. The authorities field testing is being performed from October 1986 through June 1987, first at the University of Wisconsin -- Eau Claire and then at the ten other University of Wisconsin sites.

Faxon offers direct connection to local networks

Libraries with local computer networks now can increase the speed and efficiency of access to Faxon's serial data bases via a dedicated line that provides direct communication between their host computer and the Faxon system. Through such a connection, any terminal on the local library network can access Faxon for a variety of serials control functions, without sacrificing any of the station's original uses. Faxon's LINX system supports collection development, acquisitions, claiming, cataloging, reference, and interlibrary loan. A dedicated point-to-point 9600 baud line to Faxon can reduce telecommunication and equipment costs for libraries using LINX services.

Six of the nine state university libraries in Florida are using LINX via a single drop. The remaining three libraries have an option to join and with each addition, the cost per institution decreases. Faxon also has completed point-to-point connections with a variety of other library systems, including corporate, university, and regional networks.

[Contact: Regional Faxon sales representatives or Faye Williamson, F. W. Faxon Co., 15 Southwest Park, Westwood, MA 02090; (800) 225-6055.]

Microsoft knits CD-ROM into MS-DOS

Microsoft Corp. has introduced CD-ROM extensions to the MS-DOS operating system. These extensions allow any computer running MS-DOS 3.1 or 3.2 and connected to a CD-ROM drive to read data from any CD-ROM disk formatted in the recently issued High Sierra CD-ROM file format. The extensions permit access to 550 MB of a CD-ROM disk, breaking the 32 MB file size limit normally operating in MS-DOS. The CD-ROM drive, however, appears as any other disk drive to the user and the applications software.

The High Sierra format is a logical file format that allows CD-ROM disks to be created in a single format and read by different computers using different operating systems and different CD-ROM drives. The format also is supported by Apple Computer Inc., Digital Equipment Corp., Hitachi America Ltd., North American Philips, Sony, 3M Company, TMS Inc., Reference Technology Inc., and VideoTools, among other companies.

Companies that are planning to release or who already have released CD- ROM products compatible with the Microsoft MS-DOS CD-ROM extensions include Reference Technology, TMS, KnowledgeSet, VideoTools, and Publishers Data Service Corp.

The Microsoft CD-ROM extensions will be available through CD-ROM disk drive manufacturers on an OEM basis and will be licensed separately from the rest of the operating system. The extensions consist of a hardware-independent program, a hardware-dependent device driver, which will be customized by the disk drive manufacturer or OEM, and documentation.

---

Maintained by Marshall Breeding

Copyright 1994-2024