The 2007 Open Repositories conference was in January. The distant past in internet time. But I've been taking my time to think about the conference in the context of many other internet and web-based activities and I find myself with a few open-ended questions.
(1) Can open repository (OR) software be a commodity?
First, it seems clear that more and more ORs will be used to build institutional repositories (IRs) in more and more institutions: academic, commercial, and government. Open source is hot. But widespread deployment puts pressure on open source software to be more like a commodity; viz., easy to acquire, own, and operate. And to the extent that open source software (OSS) can have these characteristics, the more it will be adopted.
In contrast to ease of acquisition, OSS promises and promotes ease of customization, and no contractual barriers to multiple use, re-use, modification, and sharing (this varies with each software package). And to the extent that a particular OSS package is customized and adapted to context, it becomes less easy to own, operate, and more unique. The mutability of software makes it hard to substitute one package for another, and makes it less like most commodity items we are familiar with; e.g., light bulbs, razor blades, printer paper, even automobiles. Computer systems aren't that easy to replace; computer applications are even less so. It's possible that the growth and spread of digital repositories is one of the first examples of widespread use of OSS. We are just beginning to learn from these experiences.
(2) Are free & open source software (FOSS) systems sustainable over the long-term?
Second, the widespread adoption of ORs also raises questions and answers regarding sustainability. Most commodity products that support the day-to-day work of schools, businesses, governments, and non-governmental organizations are provided by private businesses. No organization expects to get their physical infrastructure for free. But FOSS raises a question of whether, and how, an OSS package will be sustained over the long term (after the life of the original leaders). There's no reason to necessarily doubt the longevity of FOSS systems; many have been available for years. But there's no guarantee that they necessarily will last, just as there's no guarantee that a proprietary software product will have a long life. Of the three packages highlighted at this conference, both DSpace and Fedora are moving to non-profit organization and community-based models. EPrints has a membership model that provides a range of training, support, consulting, and hosting services.
(3) How will IR quality be assessed?
Third, sustaining a network of ORs and IRs also requires more than continuing software development, maintenance, and support. Code, standards, and protocols are necessary to support content discovery and sharing, but they are not sufficient. Interoperability is a sociotechnical attribute, and the usefulness of information in IRs also depends on sharing standards for assessment and certification of repository contents and management practices. None of this is really new (cf., the Research Libraries Group (RLG) Audit Checklist for Certifying Digital Repositories, but OR2007 featured several presentations discussing these issues and proposed solutions. In particular, MacKenzie Smith and Reagan Moore reported on the U.S. National Archives & Records Administration (NARA) PLEDGE (PoLicy Enforcement in Data Grid Environments) project that is working on ways compare policies and practices of independent and distributed institutional repositories. If the NARA/RLG Trusted Digital Repositories efforts are to pay off, then finding measurable ways to assess IRs are required. Challenges include encoding policy statements in machine usable form, and finding measurable ways to confirm collection assessment statements from IR managers. But we need to start somewhere.
It is of interest to note that there is a recently started activity that is attempting to produce an International Standards Organization (ISO) standard on which a full audit and certification of digital repositories can be based. The results of these international efforts at standardizing digital repository policies, processes, and practices are crucial components for interoperability and for sustainability.