These systems could soon offer a viable alternative to commercial products

One of the trends that I follow is the adoption of open source integrated library systems (ILS). In this column, I'll review the progress of three ongoing open source ILS projects and talk about what they mean to library automation.

Open Source in a Nutshell

Many readers likely know that open source software can be used freely without having to pay license fees to its developers. Most exists under a standard license agreement that defines the terms of use. The most common is the General Public License (GPL), which specifies that the software can be used, modified, and distributed for free. Under a GPL, the software can be changed and enhanced, but the new version must also be released under the same terms.

With open source software, the underlying source code must be made available along with the binary version that actually runs on a computer. This contrasts with the standard, commercial model of software distribution in which the source code remains the developer's closely guarded private property. Releasing source code reveals all the details of an application's inner workings. In the open source arena, this facilitates collaborative development. In the commercial arena, releasing source code can be a fundamental contradiction to basic business principles.

Koha

Koha (http://www.koha.org), which is generally considered to be the first open source library automation system, originated in 1999. The Horowhenua Library Trust (HLT), a New Zealand consortium, was in need of a new automation system. It reviewed the market and concluded that the available offerings were either expensive or lacked the desired features. HLT commissioned computer consulting firm Kapito Communications to develop a Web-based system for use in its libraries. This was a bold move. Even more remarkable, HLT determined that the new software would be made available through the open source model. This would allow other libraries to use it and make further improvements.

Koha is designed to work with a minimum of hardware resources. It runs on the Linux operating system in conjunction with the Apache Web server, uses the popular MySQL open source database management system, and is written in Perl.

Koha has been attracting considerable interest. The first system developed by Kapito was relatively simple compared to those available in the commercial arena. The features that are taken for granted in the commercial products—support for MARC record import and export, Z39.50 client and server modules, CIP or NCIP support, and authority control—were not included. A group of volunteer programmers has been working on extending Koha's capabilities to include these essential features and others. Efforts are also underway to translate the system into several languages.

So far, only a handful of libraries—including five school libraries in the Coast Mountains School District in British Columbia, Canada—is using Koha. To the best of my knowledge, no U.S. libraries had adopted it until August, when Ohio's Nelsonville Public Library (NPL) announced its plans to implement Koha. NPL is a relatively small system that consists of a main facility, six branches, and a bookmobile. It serves 36,000 active borrowers throughout Athens County with a combined collection of about 250,000 items. NPL plans to migrate from its commercially developed Spydus system to Koha by next summer.

While NPL has confidence in Koha's potential, the library requires features that the system doesn't have. So before NPL can implement Koha, these features must be added. To this end, the library has issued a request for proposal for the development of MARC support in Koha. The specific task described in the request involves adding the ability to store and retrieve records in MARC21 format. While NPL isn't prepared to fully fund all new development in Koha, it does want to offer some financial incentives to support specific enhancements. Should NPL be successful in transitioning to Koha, it would be an important step in the use of open source software in libraries.

LearningAccess ILS

Seattle's Learning Access Institute, which was founded last year, offers an open source automation system called the LearningAccess ILS (previously known as OpenBook). The organization describes itself as "a nonprofit dedicated to developing low-cost strategies to increase digital access." The development of LearningAccess ILS was sponsored by the now defunct Technology Resource Foundation (TRF). Willem Scholten, the executive director of Learning Access Institute, was also associated with TRF.

According to the organization's Web site, the institute's main goals are to develop and provide "tools for reliable information access and literacy to rural and underserved communities." The LearningAccess ILS was designed to deliver automation solutions to libraries that generally do not have the financial resources or personnel to purchase and implement a commercial system. In addition to supplying software that can be run on low-cost hardware, LearningAccess ILS assists the library in the retrospective conversion of its collection into MARC records.

Although the LearningAccess ILS tracesits conceptual beginnings to Koha, it's an entirely different implementation. This system from its very origin was designed to be multilingual. It's based on MARC21 format for bibliographic, holdings, and authority records. English, Spanish, French, Russian, and Swahili versions are currently available.

The technical components that underlie the LearningAccess ILS are similar to Koha's. The system runs on Linux, but can easily be ported to Windows NT/2000. It uses the Apache Web server and relies on MySQL as its database engine. The user interface modules are written in PHP, a Web-oriented programming language. The majority of the remaining system programming is done in Perl.

The LearningAccess ILS offers three major modules: OPAC, circulation, and cataloging. A full-fledged acquisitions module is planned for a future release.

The Learning Access Institute voices strong support for the open source development model. The LearningAccess ILS was built using only open source components. The organization plans to make the system itself available through a General Public License. Use of the software is currently limited to the initial set of libraries that are working directly with Learning Access Institute. The organization will release its system for unrestricted download once it's reached a satisfactory level of stability and maturity.

While its availability is limited, the LearningAccess ILS shows great promise. It will be interesting to monitor its development as it expands in use beyond the initial pilot libraries. To keep abreast of its developments, see the Web site at http://www.learningaccess.org/website/techdev/ils.php.

Avanti

Another automation effort that's underway is the Avanti MicroLCS. This project, begun in 1998, is essentially the work of a single individual, Peter Schlumpf. The system targets institutions by size, but is designed with enough flexibility to be adopted by any type of library. It is written in Java and will run on any operating system. Rather than making use of one of the existing databases, such as MySQL, the Avanti LCS includes its own database management system. Completion of the project's first version is expected late this year.

The Avanti MicroLCS is not in use at any library, and its source code has not yet been released publicly, pending further development. To monitor this project, see http://www.avantilibrarysystems.com.

Perspective

As I consider the library automation arena from a broad perspective, the open source systems such as the three mentioned above are but a small blip on the radar. Compared to the thousands of libraries that acquire automation systems from commercial vendors each year, the handful that use open source systems cannot yet be noted as a trend. The development and support of automation software is a half-billion dollar a year industry. Though not growing at an astronomical rate, it's generally considered healthy and competitive. Yet many libraries feel dissatisfied with the status quo. Commercial automation systems are expensive, and many libraries want to be free of the initial purchase costs and ongoing maintenance fees associated with them.

Support is a significant impediment to the implementation of open source software. The software license fees represent only one component of the overall costs that a library bears in its automation effort. The personnel required for the initial implementation and system administration is a major consideration. UNIX systems administrators continue to be in short supply and command high salaries relative to other library staff members.

Another major challenge is the ever-rising expectations that are placed on automation systems. In the commercial arena, the functionality of the basic ILS is very much a given. All the systems offer full functionality in the core library modules: online catalog, circulation, cataloging, acquisitions, and serials. Most current development lies in expanding the ILS beyond the traditional box. New development focuses on integrating content beyond traditional MARC records by bringing in book jacket images, tables of contents, abstracts, and reviews. Libraries demand increased patron-initiated Web services such as online book renewal, fine payments, interlibrary loan requests, integrated assistance through live chat, voice over IP, and video. Most of all, library systems must deliver full-fledged content, not just be finding-aids for physical collections. Although much of that content is currently in the form of text, there is an increasing amount of images, digital sound, and video. While the open source projects struggle to supply the basic features, the overall landscape is shifting rapidly.

The open source model of software development can provide an attractive alternative for libraries, if only its ideals can be realized. It's fair to say that in their current state, the open source automation systems offer only promise and potential and are not yet a viable option for a run-of-the-mill library. Even for small libraries that might be satisfied with the capabilities of the open source systems, the technical implementation and difficulty in securing ongoing support remain a challenge.

Yet, this is a trend to watch carefully. Things could change very quickly. The capabilities of the open source systems could soon surpass (or may have already) the features of some of the commercial systems that target small libraries. As libraries consider alternatives when selecting an ILS, open source systems should not be ignored. But they should also be evaluated for their own merits in features, proven reliability, support, and vision.

[Author's Note: In previous columns I've mentioned that I maintain a Web site called Library Technology Guides: Key Resources in the Field of Library Automation. The site's new URL is http://www.librarytechnology.org.]

Marshall Breeding is the technology analyst at Vanderbilt University's Heard Library and a writer and speaker on library technology issues. His e-mail address is breeding@library.vanderbilt.edu.