Increasingly, systems and tools that today’s libraries use to acquire, organize, and provide access to resources interoperate with one another. Some of those are proprietary, while others are open source. Open source software differs from proprietary software in that its license allows people to view the source code and freely use, modify, and share it. By contrast, proprietary software products are licensed for use to only paid customers, and neither paid customers nor the general public can view the source code.

Open source systems and tools were introduced to libraries more than 2 decades ago. For example, oss4lib (oss4lib.sourceforge.net/about.php), a mailing listserv dedicated for this purpose, started in 1999. These days, open source systems and tools for libraries come in many types. Open source ILSs—such as Koha (koha-community.org) and Evergreen (evergreen-ils.org)—have been implemented in many public libraries. Blacklight (projectblacklight.org) and VuFind (vufind.org/vufind) offer open source options for the discovery interface. The adoption of open source digital repository systems—such as DSpace (duraspace.org/dspace), Fedora (duraspace.org/fedora), and Samvera (formerly Hydra; samvera.org)—has been steadily growing. There are also ArchivesSpace (archivesspace.org) and CollectionSpace (collectionspace.org), open source systems for managing archival and museum collections.

Benefits of Open Technologies and Systems

Open source software is often pitched to libraries as a cost-saving measure. But the intent and the goal of the open source movement go far beyond simply providing free software. What underlies the open source movement is a particular set of beliefs about software development, where community takes a center stage. In an open source software project, a community of developers, users, and other involved parties are all expected to actively and openly contribute to the ongoing process of software development, testing, and maintenance. By making the source code available to everyone, open source software enables full participation by all who are interested and willing.

This means that in open source software projects, end users—whether they are individuals or within an institution—can play an active role in feature requests, customizations, bug fixes, user interface design, and documentations. In fact, these are the areas in which libraries often experience frustration and dissatisfaction as the paid customers of proprietary software products. A feature request for proprietary software products can take years to be fulfilled, if it is fulfilled at all. Ways to customize features tend to be limited. Bugs are often discovered only after implementation and rarely fixed promptly. The user interface can be non-intuitive and confusing. Nevertheless, libraries have to wait for a vendor to take action to resolve these issues. Switching to a different product is considered to be the last resort, since it often entails an arduous migration process.

Open source software can offer an alternative to vendor lock-in and allow libraries to take an active role in improving and customizing the software product. Imagine that libraries take the same amount of time, effort, and money that they spend on unsatisfactory proprietary software and, instead, invest in supporting and improving open source software. Also, at a higher level, the open source movement and libraries share a commitment to user-centered design. Both open source projects and libraries strive to offer systems and tools that are user-friendly and intuitive to use. For this, open source software projects solicit feedback and suggestions from users and try to promptly integrate them into the next version of the code.

By contrast, with proprietary software companies, user-centered design and responsive product improvement are not necessarily the highest priorities. Due to the unavailability of the source code, the extent to which users and customers can participate in the customization and improvement of proprietary software is significantly limited. Libraries have little influence on the long-term direction for the development of those products. Software systems and tools play the role of critical infrastructure in today’s library services and operation, but those systems and tools may evolve along the path that libraries deem to be undesirable or even harmful. This is highly problematic for libraries.

A Tale of TWO Library Browser Extensions

Open scholarship aims at producing transparent and accessible knowledge by sharing research outputs in various types for access and reuse by others from that sharing. It is easy to assume that open technologies that matter to open scholarship would be only large-scale frameworks and systems. In reality, however, many small tools, systems, and services work together to enable and support open scholarship. For example, you may think of a browser extension as a trivial tool. But a browser extension that facilitates fast-and-easy access to scholarly resources plays an important role in maintaining an open and effective scholarly environment. To further highlight the distinct characteristics of open source versus proprietary software, let me use LibX and Lean Library as examples.

LibX, an open source browser extension developed by Annette Bailey and Godmar Back at Virginia Tech in 2005, was widely used and promoted by libraries. Numerous libraries around the world customized LibX to create their own editions. In addition, librarians used LibX as a teaching tool to cover the basics of how to search for and access scholarly sources through the library. LibX was highly popular starting with its first release and quickly became a de facto standard browser extension for libraries. Its original developers and contributors continued to release updated versions to accommodate changes in web browsers and library systems for 12 years. Since its last release in 2017, LibX is no longer supported.

LibX was designed to aid scholars and researchers in quickly finding and accessing scholarly resources within their libraries inside a web browser without having to visit the library website. It worked with most major ILS and openURL link resolver solutions. LibX was fully customizable to fit the needs of individual libraries. Its website offered a user-friendly tool called “edition builder,” which allowed individual library staff members to customize, test, and publish different editions of LibX for their own library patrons.

When clicked, LibX presented a search box where you could search various systems, such as the library catalog, a list of databases and e-journals to which the library subscribes, course reserves, and so on. LibX turned unique identifiers—DOIs, ISBNs, ISSNs, and PMIDs—into links, thereby allowing users to quickly identify and access available copies through the library. It also supported reloading a publisher’s webpage through the institutional proxy server, enhanced search results in Google Scholar, and listed quick links to library services, such as interlibrary loan. In addition, on any webpage, LibX allowed a user to highlight and right-click words to bring up a context menu to search for them in preconfigured library systems. The features of LibX were well-thought-out for the use of both library patrons and library staff, because one of the two original developers of LibX, Annette Bailey, is a librarian herself.

Lean Library, a product of SAGE Publications, is designed primarily to deliver a full-text copy of an electronic article or book. Unlike LibX and most browser extensions whose use is initiated by a user clicking the browser icon, Lean Library is automatically activated when a user visits a database page or a webpage where the DOI of an electronic article or book chapter shows up. Lean Library checks the URL or the DOI against the library’s proxy config file and openURL link resolver, respectively. If a full-text copy is available through the library’s subscription, Lean Library displays a pop-up message to direct a user to the copy. Otherwise, the pop-up presents an option to gain access to the content either from another database available from the library or through the interlibrary loan service. These are marketed as Access and Alternative modules. Lean Library also has a module called Assist, which displays preconfigured specific messages for individual websites when a user visits them. The fees for licensing Lean Library are determined by the modules that a library decides to select. The pricing for academic libraries is also based upon the number of FTE (full-time equivalency) students.

For open access (OA) articles, LibX relied on the openURL link resolver only. As a product developed much later, Lean Library can connect to the APIs of the Unpaywall (unpaywall.org) and the Core (core.ac.uk) as additional sources for the availability information of OA licensed ones. Some of the attractive features of LibX are not present in Lean Library, such as the search box for various library systems to search for library resources while browsing the internet; the quick links for users to discover relevant library services; and the magic link feature of turning unique identifiers, including DOIs, ISBNs, ISSNs, and PMIDs, into links. LibX and Lean Library both enhance the search results in Google Scholar and include the context search feature that allows a user to highlight and right-click words to bring up a context menu to search for it in the library catalog.

Some TakeAways

LibX and Lean Library were developed at different times. The first version of LibX was released in 2005, and its design didn’t change much. Lean Library appeared much more recently. LibX aimed to lead users to actively search and discover library resources and services. By contrast, Lean Library has a narrower focus of simply delivering full-text copies of electronic articles to users when available. Users have a much more passive role in Lean Library compared to LibX. This is important because it bears on how libraries want to serve library patrons through a particular tool such as a browser extension. For example, LibX was used as a popular pedagogical tool among librarians. It would be difficult for a product like Lean Library to be used for such a purpose. 

LibX and Lean Library took different approaches to configuration and maintenance. LibX’s integration with various library systems was based upon standards such as the openURL and the http Get request. This minimized the additional setup and ongoing maintenance work needed for an individual edition. By contrast, Lean Library’s Access module requires the library’s proxy config log to be ingested on a regular basis in order to extract the domain information of subscribed databases. Since the proxy config file is not intended to be used for this purpose, individual libraries must spend time and effort to remove many false positives. Its Assist module also requires a custom message to be manually created for each website. These makes Lean Library’s initial setup and ongoing maintenance more time-consuming to the library staff.

Also, libraries that license Lean Library are unlikely to form a community similar to that of LibX. Even though the LibX project did not succeed in generating a user and contributor community that could sustain it in a reliable manner, a community of the individual LibX maintainers actively shared information about different configurations and customizations. Because LibX was open source, library staff were easily able to look at, copy, revise, and test the configuration and other settings of other libraries’ LibX editions. This is not a possibility for commercial products like Lean Library, whose code is not available for viewing or customization.

With the loss of LibX and the emergence of commercial browser extensions such as Lean Library, it is important to note that the library community, which could have had two distinct types of library browser extensions with an emphasis on different aspects of library services and different roles of users (active vs. passive), lost one of those options. Open source systems and tools, of which LibX is an example, afford libraries the opportunity to directly impact and shape their future evolution in ways that closely reflect libraries’ own educational concerns and future vision on open scholarship and open knowledge. The library community as a collective failed to act to ensure the longevity and sustainability of LibX, an excellent open source tool. It did not initiate any movement to invest in this project through funding or code contribution and maintenance. Librarians and library staff enthusiastically adopted and customized LibX, but they did not manage to organize themselves into an active community with a systematic way to participate in and grow the project. When LibX went dark, many libraries that had been heavily relying on it were caught off guard. Vendors quickly developed proprietary browser extensions for this newly opened market.

Investing in Open Technologies as a Spending Priority

If libraries want to actively shape their path toward better library services and operation over the long term, they should make collective efforts to invest in open technologies and open source projects. Only then will libraries not be in danger of losing a valuable tool like LibX. The adoption of open source software has been relatively slow due to a number of reasons. Libraries favor the user-centered and community-driven spirit of the open source software projects over the closed vendor systems. But they also need quick and reliable support for the open source systems and tools along with assurance about their longevity and sustainability. In addition, small and mid-sized libraries face the lack of sufficient in-house IT and computer programming skills, which are often required to install, customize, maintain, and contribute to library open source software.

Fortunately, library open source projects have been getting more stable in recent years. Many of them have established formal community governance structures and started to operate with long-term business models. More companies specializing in supporting open source software have also sprung up. Contracting with those companies offers an attractive alternative for libraries that are interested in adopting open source software but lack necessary in-house IT expertise.

A concrete movement for libraries to invest in open technologies as a spending priority is also surfacing. The Invest in Open Infrastructure (IOI) initiative (investinopen.org) began in 2018 with the aim of improving funding and resourcing for open technologies and systems that support research and scholarship. In order to fully support and sustain the practice of open scholarship, libraries themselves must invest in the critical components of the open scholarly and research environment to ensure that they grow and prosper. For it is only in such an environment that libraries will be able to see knowledge be easily accessed, openly shared, accurately reproduced, and appropriately reused, just as they desire.