Diane R. Tebbetts
University of New Hampshire Library
Durham, NH 03824-3592, USA
Abstract: In the world of library automation, it is evident that
change is a way of life. Upgrading systems, switching to new systems, or
interfacing modules from different systems occur with increasing frequency.
As a library is faced with this situation, it becomes essential to evaluate
systems from a different perspective. Some of the factors to consider in
migrating to a new system include: the vendor's experience with migration
and data transfer, the system's capability for expansion, the vendor's
research and development program, the networking capability of the system,
the flexibility of the system, the reporting features of the system, and
the stability of the vendor.
Library automation is a dynamic, fast-paced process, and change in hardware, software, and total systems is inevitable. Two in-depth studies (1985 and 1990) of the nine ARL (Association of Research Libraries) libraries in New England revealed the extent of change that had occurred in this five year span. Upgrading systems, migrating to new systems, interfacing modules from different systems occur with increasing frequency. Regardless of whether libraries have been automated for several years or a few, whether they have integrated systems or interfaced modules, whether they are fully automated or partially implemented systems; they still will need to make changes. Increasingly, libraries are forced to migrate to totally new systems. Planning for a second system is different from planning for a first system. It is necessary to consider new criteria and evaluate systems from a different perspective. This paper examines the reasons for change and factors to consider in migrating to a new system.
The high rate of change in library automation is one of the most significant findings of this researcher's study of New England ARL libraries. Between the original research in 1985 and the follow-up study in 1990, several of these libraries had undergone major changes in their systems. In 1985, 5 of the 9 libraries had purchased or were in the process of purchasing integrated online library systems. In 1990, 4 of the 5 had either to change systems or to add modules from other systems. The library that had a circulation system was considering developing a distributed inter-faced system with a CD online catalog. Of the 2 libraries with in-house systems in 1985, one had chosen a commercial system while the other had continued to develop its own system. One library that did not have an automated system in 1985 had chosen an integrated system in 1990. Only 2 of the libraries were on the same course in 1990 as they had been in 1985. One of these was an inte-grated system with the same vendor and the other was an in-house system.
This high rate of change is not unique to the New England ARL libraries. In fact, other studies indicate the same findings. For instance, a national survey conducted by Cahners Publishing for Library Journal reports that more than 60 percent of the systems reported had been installed in the last five years and even so 58.5 percent of the public libraries and 40 percent of the college and university libraries had already upgraded their systems (Berry, 1989).
3. REASONS FOR CHANGE
There are several reasons for the high rate of change in library automation. To a large extent, the library automation field reflects the volatility and high rate of change in the computer market-place as a whole. Some of the factors impacting on library automation are:
3.1. New Technological Developments
A major factor in the high rate of change in library automation is the fast-paced development in the field of computerization (Figure 1). Library automation vendors have to keep up with the new develop-ments in hardware and software design. In addition, they have to provide the modules needed in the library workflow. Some vendors do not offer complete systems and they have to upgrade their systems to include the needed subsystems. For instance, some vendors do not have serials control subsystems or acquisitions subsystems. Then, the libraries have to add modules from other systems, wait for their vendors to develop the needed modules, or buy completely new systems.
Library automation is a continually evolving process. Libraries find they have to make changes to provide the necessary services. Added software to implement new modules, upgraded systems or entirely new systems require new more powerful hardware. New hardware is smaller and more powerful. It is becoming a difficult decision to determine what type of hardware to purchase. A trend in computerization in general and library automation in particular is "down-sizing." "Loosely defined, downsizing usually refers to shifting applications from a mainframe to a smaller platform, such as a midrange system, a UNIX server or a PC local-area network" (Radding, 1992). Frank Bridge in the "Automated System Marketplace 1992" reports the same trend: "As we continue to examine the prevailing trends in library automation, the major emerging theme is that library automated system downsizing has continued and accelerated" (Bridge, April 1, 1992).
This trend to smaller, more powerful computers, combined with open systems, has complicated the library automation choices. "An open system is any system that can communicate with another open system using well-defined national and international standards" (Hinnebusch, 1991). Hard-ware and software used to be sold together as a package. With open systems this is no longer necessary or even advantageous. However, it does make the choices more complicated and the changes in systems more dramatic.
With these developments in computer technology it is obvious that library automation is changing rapidly to take advantage of new hardware and sophisticated software to improve library processes.
Networks from local area networks (LANs) to campus networks, to state and regional networks to the high speed networks such as Internet and NREN (National Research and Education Net-work) have had a major impact on library automation. The LANs have facilitated the development of downsizing by allowing distributed databases to operate transparently to the user thereby simulating a single large system. Whether or not a library chooses this type of system, it will be essential for the library to connect easily to local area networks. This may require changes in the library's system including upgrading or switching to a new system.
In addition to connecting to LANs, it is essential for academic libraries to have their systems available on their campus networks. Faculty and students want and will demand access to the library's online system. The New England surveys indicated the dramatic changes in network availability between 1985 and 1990. In 1985, half of the ARL libraries were connected to a cam-pus network while all were connected in 1990. The need to access the online catalog, to provide additional databases, and to supply documents on demand requires upgraded hardware and software or entirely new systems.
In addition to connectivity to the campus networks, connectivity to the high-speed national networks is becoming essential. In the New England surveys, 1985 revealed no libraries connected to the Internet while all but two were connected in 1990 and half had their online catalogs availa-ble. Currently, the Internet connection is even more important, with discussions on the Internet revealing that libraries are providing Internet public-access workstations for their patrons. Again, changes in hardware and software may be necessary to provide this access.
Therefore, the fast-paced developments in networks and telecommunications are requiring libraries to make changes in hardware and software to provide the latest networking capabilities to their patrons.
3.3. Expansion of Databases
It became clear in these New England studies that libraries are not only providing access to their online catalogs but they are also adding other databases to their online systems, including periodical indexes, full-text databases, and locally-produced databases. Everything from Current Contents and Medline to Shakespeare's plays were being added. These databases very often require expanded hardware facilities and new or upgraded software. This means that the library has to change or expand systems to supply the needed capabilities and capacity. Needs in library automation change so rapidly that it is essential for the library and its system to have the flexibility to accommodate new demands.
These added requirements help to create the dynamic, changing, fast-paced environment in which library automation operates.
3.4. Vendor instability
Perhaps the single, most-important factor contributing to the fast pace of change in library automation is the instability among library automation vendors. Of the 5 library system vendors that had installed or were installing the integrated systems in the 1985 survey, 4 had either gone out of business or were experiencing financial difficulties in the follow-up survey in 1990. Since that time, the fifth vendor has been bought by a different company. Therefore, all 5 vendors have experienced major changes in a 7-year period.
The experience in New England reflects the situation nationally and in Canada. A major finding of the fourth annual survey of Canadian libraries indicates that "The result is that the large integrated library systems dominating the market at the end of the eighties are totally different from those that dominated it at the beginning. . ." (Merilees, 1990). This vendor volatility com-bined with the fast-paced nature of library automation creates a difficult situation for library managers. To keep up with increased demands and a changing environment, libraries need a system that is being upgraded and meeting current needs. When the system is no longer viable or functioning effectively, it becomes mandatory that the Library migrate to a new system.
4. MIGRATION TO A NEW SYSTEM
After the library decides it has to have a new system, it has to go through the process of evalua-ting and selecting a new system. However, it is quite different the second time around. In some ways it is easier, but in others it is more difficult. This time, of course, the library will have an experienced, knowledgeable staff, a machine-readable database, and a barcoded collection. On the other hand, the critical questions will be:
• Will the data transfer to the new system?
• Will the barcodes be compatible with the new system?
• Will the new vendor understand the library's migration needs?
Therefore, there are several factors that need to be considered when a library is switching to a new automated system in this current climate of fast-paced technological developments, vendor instability, and expanding needs and demands. As a library is faced with this situation, it becomes essential to consider systems from a different perspective. Some of the factors to consider in migrating to a new system include:
4.1. Support for Migration
A critical factor in selecting your new system should be the experience and reliability of the proposed vendor in migration. Of particular importance will be whether this vendor has experien-ce in migrating libraries from your specific system to the proposed system. "A vendor's experience with converting your existing system is the primary predictor of project timeliness and success" (Bridge, May 15, 1992).
It is essential to try to determine whether the vendor has converted any systems. In any discus-sions, try to determine this factor. If possible, secure the names of libraries that have migrated to the system. Librarians at these installations can give first-hand, detailed knowledge of the length of time the conversion will take, data that may be problematical, precautions the library might be able to take, and preparations that would be helpful to the library and to the vendor. They can also give some indication of how responsive the vendor is to the library's needs and the amount and kind of support provided to the library staff. This will be of major importance to the successful migration from one system and vendor to another.
4.2. Adherence to Standards
As we have just discussed, the ability to migrate data from the current system to the new sys-tem is of major importance in implementing a second system. Clearly, this can be problematical because of the many kinds of data that must be transferred – MARC records, patron files, and item records to name some. There will be a much better chance of success if the new system adheres to as many standards as possible; for instance, the recognition of standard bibliographic records. If your current system has created and maintained standard USMARC records and your new system will be able to read them and load them, you should be able to transfer your biblio-graphic records relatively successfully. If your current system has done anything unusual to your records, you may be in for some conversion difficulties. In your new system, it is essential to determine that it adheres to the standards. It is very likely that you will be migrating to another system in the not too distant future and it will be essential that standard MARC formats are employed and no unusual modifications are made to the data.
4.3. Capability of Expansion
It is obvious from this author's New England studies that libraries are adding many different kinds of databases to their online catalogs. As discussed previously, these databases include externally-developed indexes such as Medline and U.S. government publications; full-text data-bases of periodical articles, encyclopedias, and literary works; and locally-produced materials including images. To add materials such as these requires much capacity in your system. Be sure to buy a large enough system or one that is capable of expansion and or interfacing with other modules and systems.
In addition to the size of the databases, be sure to allow for enough users. As a system be-comes easier to use and more available on networks, the number of simultaneous users will in-crease. Be sure to consider the possibilities for expansion to increase access. Faculty and students will want to access the system from offices, homes, and dormitory rooms. In addition, off-campus users will want to access over the networks. Make sure your new system has the capability of expansion or you will have to change systems more quickly than anticipated.
4.4. Support of the Vendor
In selecting a new system, the support of the vendor is critical to the success of the system. Especially important is the role of the vendor in the installation of and training for the new system. Equally important, however, is the long-term support for the system. Does the vendor have an active research and development program? Are upgrades provided regularly? If the vendor is strong in this area then it is likely that the system will keep up with new technological develop-ments.
Are all the advertised modules fully functioning? Too often, vendors claim modules are avai-lable when they have not been fully tested and "debugged." How responsive is the vendor to libraries' inquiries? Very often, a check with current customers will answer some of these ques-tions. Genaway, in a 1989 Technicalities article, suggests some of the questions that might be asked of libraries using the system under consideration. This is a very good way to determine what the system is like in operation (Genaway, 1989).
4.5. Networking Capability
Of major importance in library automation is the ability to connect to all kinds of networks: campus, regional, and the high-speed national networks. This has been a fast-paced development in the world of telecommunications. In the 1985 study, no New England ARL library was con-nected to a national network, while all of the academic ARL libraries in New England were connected in 1990 and half had their online catalogs available on the Internet. In addition, all of them were connected to campus networks. This capability is essential in any new system.
To plan for the future in networking, a system that adheres to standard telecommunication protocols will have more flexibility and connectivity. The ability to query different systems and databases using a common protocol is essential to the ease of access. Therefore, systems employ-ing standard protocols will be more viable and capable of more readily connecting with other systems and accessing remote databases.
Therefore, it is important to determine what the vendor is doing in this aspect of library automation to be prepared for future developments.
4.6. Flexibility of the System
The more flexibility the new system has, the more it will be able to adapt to the changing environment. Systems that can operate on multiple platforms are becoming a very attractive alternative. Unix as an operating system for library applications is attracting interest because it can run on a wide variety of hardware from micros to minis to mainframes, and because it can handle several operations. "Unix remains compelling because it is the only operating system to offer multitasking, graphics, and cross-platform compatibility in one package" (Yager and Smith, 1992). Unix is especially interesting in the context of library automation because: "Unix has features that allow it to easily accommodate information as text, that allow several tasks to be handled together, and that allow many users to access the same information simultaneously" (Brandt, 1991).
Whether one chooses a system using Unix or another operating system, the most important consideration is its flexibility. It is likely that proprietary systems will be more limited and be less likely to adapt to future changes and new technological developments. Therefore, buy a system with as much flexibility as possible.
4.7. Production of Management and Statistical Reports
It will be very important to the library manager that the new system provide reliable and exten-sive reports. The more information that is available on the use of the system, circulation of the collection, acquisitions of materials, cost of items, and success of users, the better decisions library administrators can make. The better the information and the more data that are available, the easier it will be for librarians to adapt to the changing economic environment, to meet the new needs of users, to adapt to the changing curriculum, to exploit the new sources of information, and to incorporate new technological developments.
Therefore, it is essential that the new system provide relevant data in easy-to-read and compre-hensive reports. It will be important for this to be a major consideration in evaluating a new system.
4.8. Stability of the Vendor
Of all the factors this is probably the most important to the success of a new system. At the same time, it is probably the most difficult to determine. There will probably be no way to be absolutely certain about the viability of a vendor. However, there are some things that can be determined. Of course, the vendor should be willing to supply a current financial statement or annual report which should include information on the number of employees, size of the research and development division, number of installations, number of recent sales; in addition to current financial data such as profits, losses, and operating expenses. Of course, a good balance sheet is no guarantee that the vendor is going to be viable far into the future, but as Bridge states: "This pro-cess can provide you with valuable insight into your vendor's prospects for the near term. If the vendor cannot pass the near-term viability test, the long-term prospects become irrelevant" (Bridge, May 15, 1992).
Clearly, it is essential to have a vendor that will provide the support, training, research and development that will make the system adaptable to the changing library automation. Therefore, the more information you can find out about the vendor from data supplied by the vendor, from research reports and experts' reports, and from the experience of users of the system, the better decision you will be able to make when selecting a new system.
From the preceding discussion, it is obvious that the high rate of change in library automation means that libraries will have to upgrade systems, add modules from other systems, or purchase completely new systems. Frank Bridge estimates ". . . a library system cannot remain reliable much longer than about five years" (Bridge, May 15, 1992). In this type of environment it is essential to plan for change. It is obvious that selecting a second library system is different from purchasing a first system.
Of primary importance in the process of migrating to a new system is the transfer of data. It is essential to determine how the vendor handles the transfer, if the vendor has had experience load-ing tapes from the previous system, and if and how the new system handles MARC bibliographic records. It is important to remember that your database is the essential element in your online system. Restoring lost or corrupted data will be time consuming and costly.
In the fast-paced arena of the library automation marketplace, your choice of vendor will be critical to the success of your conversion project. Do investigate the vendor's situation as thoroughly as possible. A vendor with strong research and development capability that provides upgrades to the system will help to stretch the life-span of the system.
Finally, plan for your system to allow you as much flexibility and maneuverability
as possible. This will allow you to take advantage of new technological
developments and change with the least amount of difficulty. In your planning
always be prepared for change.
Berry, John. (September 15, 1989). "Upgrading Systems, Software & Microcomputers," Library Journal, pp. 56-59.
Brandt, D. Scott. (February 1991). "Who Can (or Should) Use Unix?" Computers In Libraries, 11 (2): 32.
Bridge, Frank R. (April 1, 1992). "Automated System Marketplace 1992," Library Journal, pp. 58-72.
Bridge, Frank R. (May 15, 1992). "System Migration: Abandoning Your Vendor," Library Journal, pp. 68, 70.
Epstein, Susan Baerg. (January, 1991). "Implementing a Second System: Some New Concerns," Library Journal, pp. 76-77.
Genaway, David C. (October 1989). "Micros, Minis, and Mainframes," Technicalities 9 (10): 4-7.
Hinnebusch, Mark. (April, 1991). "Integrated Library Systems," Academic and Library Compu-ting, 8 (4): 6-8.
Merilees, Bobbie. (June, 1990). "Integrated Library Systems in Canadian Public, Academic, and Special Libraries: Fourth Annual Survey," Canadian Library Journal, 47: 193-200.
Radding, Alan. (August 10, 1992). "Dirty Downsizing," Computerworld, pp. 65-67.
Yager, Tom & Smith, Ben. (September, 1992). "Is Unix Dead?" Byte,