EXPANDABILITY, FLEXIBILITY, COMPATIBI-LITY: KEY MANAGEMENT CONSIDERATIONS IN ACADEMIC LIBRARY AUTOMATION

EXPANDABILITY, FLEXIBILITY, COMPATIBI-LITY: KEY MANAGEMENT CONSIDERATIONS IN ACADEMIC LIBRARY AUTOMATION

Diane R. Tebbetts

University of New Hampshire Library
Durham, NH 03824-3592, USA

This research based on 1985 and 1990 survey data of the Association of Research Libraries (ARL) in New England provides insights into the direction in which library automa-tion in New England is heading. By comparing 1985 and 1990 results, it is possible to see changes in library administration and planning, as librarians strive to manage in the volatile, fast-paced environment of information technology. The nine libraries studied are:

They are the largest libraries in New England, and their experiences in automation are expected to be representative of those of other major academic libraries in other regions of the U.S. and Canada and foreshadow the developments imminent in smaller institutions.

From the experiences of these libraries, it is possible to determine some of the major trends in library automation and the strategies employed to direct the development of automated library systems.

In 1985, these libraries were involved in coordinating and unifying the automation efforts that had been proceeding in the 1960's and 1970's into a unified library system for each library. Six of the nine libraries had decided to implement automated integrated systems, two had inhouse systems, and one had not determined its future direction. In 1990, these libraries were still in the process of coordinating their internal processes into a unified system. Many had had to change systems or add new elements to their systems. Many of these changes were necessitated due to the volatility of the library system marketplace, but some resulted from the advancement of technology. In 1990, the libraries were very concerned about expanding their resources beyond the library's own system by adding externally-developed databases, by net-working to other systems, and by providing resources to their faculties and students beyond the library walls. These new directions in library automation resulted in new emphases in library decision-making. We will explore these trends further in the discussion that follows.

The author's 1985 research into library automation in the New England ARL libraries produced a comprehensive review of the state of their automation efforts up to that date. A 1990 follow-up survey of these same libraries provides a longitudinal study of library automa-tion developments. Employing the case study method, library directors and/or systems librarians were interviewed in 1985 and 1990. The open-ended interview technique was employed to elicit more information from the interviewees.

• What are the major factors in the decision-making process for library automation

From these study results, it is evident that the emphasis in library automation is changing. In 1985, the libraries were concentrating on combining the separate library functions (i.e., cataloging, acquisition, serials, etc.) into a unified library system with an online public access catalog. These systems were either integrated with one database serving as the basis for all library functions or interfaced with functions handled by separate modules but linked to provide the appearance of a single system. In 1990, these libraries were still involved in this process. Some had suffered setbacks due to vendor instability or incomplete systems and had had to switch systems or buy modules from other systems. In New England, four of the five vendors with which the libraries had contracts in 1985 were either out of business or experien-cing difficulties in 1990. A survey of integrated library systems in Canadian libraries revealed this same situation. In the early eighties the vendors dominating the market were completely different from those at the top in the late eighties (Merilees, 1990).

Therefore, the New England libraries were still in the process of consolidating their functions to produce a unified system. However, new developments in the library/information field were adding new dimensions to the automation efforts. Although the automation of the local library's functions was still extremely important, the ability to expand the resources and connect to outside databases was becoming essential.

These libraries in 1990 were involved in expanding their resources by adding externally-produced records -- for instance, Center for Research Libraries data and Government docu-ments tapes. Other libraries were adding externally-produced databases to their online cata-logs, both bibliographic and full-text. MEDLINE, Current Contents, and Shakespeare's plays are just some examples of the range and extent of these activities.

These libraries are also integrating other media into their systems. Integration of CD-ROM technology and the online catalog is an area of activity as well as tape-loading on the local systems.

In addition to expanding the local resources by adding records and databases to the indi-vidual library's system, access to other databases and other libraries' holdings via the national highspeed networks is becoming increasingly important. The ability to connect to networks, both local and national is a major factor in library automation. In the New England study in 1985, only half of the libraries were connected to a campus network and none were connected to consortium or national networks. In 1990, all the libraries were on a campus network; all but one were connected to the Internet, the national high-speed network; four had their catalogs available on the network and three were connected to a consortium network. The development of networking from 1985 to 1990 was a major factor in the development of library automation.

In this incredibly fast-paced environment where the ability to mix and match systems, to access other databases, to load tapes from outside sources, and to connect to networks are essential; it is evident that new elements must be considered in the decision-making process. From this study, it appears that some key concerns in library automation are expandability, flexibility, and compatibility.

This study shows that certain trends in current library practice are related to the needs to expand the online catalog, to intermix disparate systems, and to access networks. Keeping these needs in perspective, the interrelationship of the following trends becomes apparent.

In 1985, many of these libraries had purchased or were looking at vendor-supplied hard-ware/software packages. However, this was a risky choice because of the volatility of the library automation marketplace and the instability of the vendors. As discussed previously, this situation is reflected in the Canadian study. In 1990, these libraries were no longer looking at proprietary systems, but were concentrating on the large computer companies such as IBM and DEC. Five of the libraries had or were looking at systems that run on IBM, two had DEC equipment, one had a Data General (DG) machine, and one was interested in equip-ment running the UNIX operating system. These results correspond to those of the 1989 Library Journal survey of over 1,000 libraries which found that more than 46 percent of those libraries with systems operate on DEC or IBM equipment. The remaining equipment was spread over six to ten manufacturers with DG's share approximately three percent (Berry, 1989).

Certainly, the vendor instability is one factor in this shift to the major hardware manufacturers. Other factors include the need for compatibility of systems. With libraries finding it necessary to mix modules from different vendors, to access other systems, and to connect to networks, it is more feasible if standard equipment is being used.

The Canadian study also found that hardware decisions were a significant factor in library system selection. Therefore, a library should be very careful when selecting a system. A system that will run on different hardware and is flexible is, of course, the optimum solution. However, buying systems that run on standard equipment rather than proprietary hardware seems to be the route that many libraries are following.

From the trends evident in this study, it is obvious that libraries are in the process of expanding their resources to provide additional materials either by loading databases locally or providing access to remote databases. In addition, they are providing more access to their own online catalogs through campus and national networks. This type of expansion requires increased computer capacity. The libraries were adding bibliographic and full-text databases -- two have loaded MEDLINE, one Current Contents from the Institute for Scientific Infor-mation, and one has a full-text encyclopedia and Shakespeare's plays online. All of the acade-mic libraries are connected to their campus networks and four have their catalogs available on Internet. One library has 6,000 terminals connected to the system with 60-65 simultaneous users; another has 425-430 dedicated terminals, and still another logged 16,000 dial-in ses-sions in one year.

With all this activity, the library system must have a large capacity. It is essential that this be a major factor when considering library systems, or the library will find very quickly that it has purchased a system with inadequate capacity. The Library Journal survey found that "A more troubling part of the need to upgrade is that libraries have apparently purchased systems of either inadequate capacity, or that lag behind the current state of the art in computer techno-logy" (Berry, 1989).

Librarians must be aware that the current directions in library automation will make it necessary for them to have systems with a large capacity so they should make that a require-ment rather than opting for small systems with which to begin. A small system can end up being an expensive choice.

Of all the trends in library automation, this is probably one of the most significant. As we have seen, the changes in networking capability between the 1985 and the 1990 study were dramatic. The libraries connected to campus networks increased one hundred percent while library participation in the national networks came into being during this period. Connection to other networks such as consortium and regional networks, also showed dramatic increases during this period.

The ability to connect systems and provide access to these networks becomes a crucial factor in library decision-making. Systems that connect readily and employ standard conven-tions become the best choices in this situation. As Susan Martin states "... the need to link local systems to other local systems, bibliographic utilities, and remote databases will become critical. The most valuable task that librarians can perform is to ensure that the local systems they specify and purchase have the capability of using OSI protocols to communicate outside the institution." (Martin, 1989).

Therefore, the findings of this study confirm that telecommunications and networking are major factors in library automation and predict that they will be of increasing importance in the future.

In this fast-paced, quickly-changing environment, it is becoming very important to buy library systems that are as flexible as possible. In this study, the libraries were looking at systems that would run on a variety of hardware and that could connect easily to other systems and networks. One library purchased a system that would run on any UNIX-based equipment. The Canadian survey indicates that several vendors have re-written their systems to run under the UNIX operating system because this enables the programs to work on a variety of hard-ware. (Merilees, 1990). Another library in this study bought a system that would run on DEC or IBM. Bridge, in his survey of the automation marketplace, notes that there is a growing trend to purchase "software-only" systems which will run on existing equipment. (Bridge, 1991).

Software that will run on a variety of hardware is becoming an increasingly attractive option in this environment where flexibility and compatibility are key considerations.

In this highly interactive world where the ability to connect to campus networks and other computers is essential, standard protocols become vital. In considering an automated library

system one should keep the ability to interact with other systems and the use of standard

communication protocols as key considerations. "The choice of which protocols to adopt may seem like a technical issue, but it's one that campus administrators cannot afford to ignore. More than one campus found they had erected 'electronic curtains' rather than connective infras-tructure by not paying attention to such details ..." (Leach, 1990). Become familiar with TCP/IP, OSI, and Z39.50. Know about these standards and know whether your equipment employs standard protocols and which ones. The ability to connect to networks, link with other computers, and access remote databases will be critical in the evolving information world.

As much as possible it is essential to work with standard records for bibliographic data. In this fast-paced, changing environment, the record becomes a constant element. When you have to change to a new system or link with another computer, your record should be a stan-dard one. This means that you should not customize records for one system. Make sure that your system uses standard bibliographic MARC formats. Don't allow fields to be stripped from your records. If you want to do this for a specific application, make sure your archival records have retained the full record. Replacing what you have deleted is extremely costly. As Clifford Lynch states: "We understand the data elements, at least for bibliographic data, if not quite so well for abstracting and indexing data. We have the MARC formats, so we have

someplace to start when we talk about how to search this information and how to move it from place to place in a comprehensible transfer format." (Lynch, 1990). A standard record becomes an element to rely on in a world where other elements are changing rapidly. A stan-dard record allows for flexibility in other areas, such as hardware and software.

In this highly volatile world where vendors change rapidly, hardware becomes outdated quickly and software is updated frequently; the need to manage the library automated system must be done locally. It is no longer possible to rely on the expertise of the vendors. A library must understand its own situation and manage the system to fulfill its own needs. In addition, the ability to add records from other sources, to connect to local and national networks, to interface disparate systems and unlike equipment requires specialized expertise. Therefore, it becomes essential to have personnel in the library acquainted with these issues.

In this study, it became obvious that the ARL libraries in New England were responding to these needs by expanding their systems departments. All of the libraries saw an increase in systems personnel between the 1985 and 1990 surveys. One library's system staff increased by 300% while another library completely reorganized its system office. The average size of the systems office was four persons -- director, programmer, operator, and documentation specialist. It is essential to remember that these increasingly complex systems require increased systems staffing.

In this automation environment where systems change quickly, hardware is superseded, and software upgraded regularly, the database along with the individual record becomes extremely important. The database has to be maintained, standards applied, and disparate records integrated into a coherent whole. This means that someone has to supervise the database and maintain it.

In this study, it became apparent that the ARL Libraries were also recognizing this fact. In some of the libraries, the supervision of the database was taken care of by technical services staff, while in other libraries systems staff had responsibility for the database. One library had created the position of database librarian whereas another library had established a bibliographic control unit.

There are different models but it is essential for someone to be in charge of maintaining the database. With records being added from various sources of which the library's cataloging department is only one, this supervision is vital to the integrity of the system. The database becomes the essential ingredient in the library's automated system and must be built and maintained with care.

Automated library systems do not operate in isolation. They are evolving, interconnected systems that are changing constantly. To be useful, they must interact with other computers, both on and off campus; they must connect to local, regional, and national networks; they must be compatible with other campus equipment; and they must be accessible to users. This type of linking and interfacing is not easy. It takes the expertise of library and computer personnel working together.

This study revealed that the New England ARL libraries were working closely with their campus computer services. In some universities, the relationship was very close, with the library and computer services sharing responsibilities. For instance, in some cases the computers are actually located in the computer services facilities. In other cases, the computer services acts in an advisory capacity to the library. In still other situations, the library maintains some functions such as programming and front-line support while computer services provides operation management and back-up support. In all cases, it is becoming essential for the library to have close relations with the campus computer services. The computers must communicate with one another so the people must talk to each other.

As we have seen, the world of library automation is a vital, changing environment. Systems are changing rapidly. The ability to plan ahead and be prepared to change systems is critical. In this study, one director stated that automation is a continuous process while another library administrator indicated that while she was purchasing one system she was planning for the next.

It is important to remember that it is not the hardware and software that are critical but the records and the database. As Karl Beiser states: "Remember, you will never buy your library's last computer. Sometime in the future whatever you are doing now will be able to be done so much better with new equipment that you will buy something else." (Beiser, 1991).

Therefore, it is essential to plan for the next system. Try not to make decisions that will limit your flexibility. Do not sacrifice your database to meet equipment needs. Equipment changes but your database remains. Making decisions with this in mind will help avoid costly corrections in the future.

From this study, it is clear that library automation is a complex, dynamic, and interrelated process. Library systems are not static, isolated structures but rather evolving, linked systems. Unquestionably, new records and databases are being added; connections to networks and other computers are being developed; access by local and external users is being required. In this environment, expandability, flexibility, and compatibility are key considerations.

Keeping these requirements in mind, decisions about hardware and software become relative and changeable, whereas the quality of the records and the database becomes a long-term concern. The adherence to standards and supervision of the database become increasingly important as factors in maintaining flexibility and compatibility. Planning for the future requires an understanding of the automation process -- the need to recognize the ephemeral nature of equipment and the ability to make decisions that will allow the flexibility to change.

Beiser, Karl, "Library technology through a wide-angle lens," Wilson Library Bulletin 65 (9): 48-50, 157 (1991).

Berry, John, "Upgrading systems, software & microcomputers," Library Journal pp. 56-59 (September 15, 1989).

Bridge, Frank R., "Automated system marketplace 1991," Library Journal pp. 50-62 (April 1, 1991).

Leach, Ronald G., "Campus networks: Leveraging investments in automation, faculty, and libraries," Library Issues 11:4 (September 1990).

Lynch, Clifford A., "National networking and the University of California: Goals, objectives and future developments," Presentation to OCLC Users Council, February 5, 1990. (unpublished)

Martin, Susan, "Information technology and libraries: Toward the Year 2000," College & Research Libraries 50 (4): 397-405 (1989).

Merilees, Bobbie, "Integrated library systems in Canadian public, academic and special libraries: Fourth annual survey," Canadian Library Journal, 47: 193-200 (June 1990).