Socioeconomic issues associated with the delivery of online content and services are as critical to success as sound technological implementation, if indeed not more so. ACM distinguished itself in its advanced planning for its digital library by paying close attention to three such issues: changing patterns in scholarly behavior; the functioning of copyright law in a networked environment; and the development of a business model.
There has been an explosion in the amount of research formally published. There is a high degree of specialization, leading to more journals, each serving smaller communities and simultaneously creating the need for a more general type of publication that translates specialized results into more commonly accessible language, teasing out possible implications for practical development. Many of the production values of print still outweigh their online equivalents. But the great power of computer searching and linking over manual indexes and references is clear, perhaps not within a single journal but most assuredly across a digital library. The most valuable commodity of the researcher is his or her time, and a most pressing need is keeping up. ACM concluded that a well-indexed, searchable database of articles had a greater chance of helping scholars than simply providing subscriptions to online journals. Unlike many other publishers, ACM made a strategic initial decision to orient its online development around a digital library rather than electronic journals.
Two years before releasing its digital library, ACM worked on an approach to managing intellectual property in the networked environment. The ACM Interim Copyright and Permissions Policy, published in the April 1995 issue of Communications,1 was considered a landmark intellectual property policy soon imitated by many other publishers. It articulated appropriate uses in light of the massive republication possibilities of the new medium. It treated the relationship of traditional references to incorporative links. It extended ACM’s traditionally liberal copyright transfer policy that gave authors explicit rights to reuse their own works and generally broad educational uses, well beyond the fair use provisions of the copyright law itself. The policy was developed to a large degree in the absence of relevant case law and prior to the new Digital Millennium Copyright Act.
An essential assumption of the ACM policy was (and is) the community would honor the Permissions Policy even if the notion of "personal use" was left somewhat vague. It was further assumed the cost of technological enforcement against illegal downstream uses outweighed the benefit, particularly if it meant encumbering legitimate usage with unfriendly obstacles. Rare cases of systemic infringement could be spotted and dealt with after the fact.
A viable business plan is a requirement for any long-term, self-sustaining digital library. Establishing such a plan for ACM required a balancing act. Conversely, ACM had a tradition (and hence both an expectation and financial need) that publications generate surplus revenue to support other activities of the Association. On the other hand, there was a strong Internet new frontier ethos that "information seeks to be free," coupled with a backlash against escalating prices for scientific and technical journal subscriptions. The tension between free-and-easy dissemination and the demand for revenue is certainly not new in association publishing, but it is exacerbated in the networked environment.
ACM took a conservative approach: aiming at net revenue neutrality in shifting from print subscriptions to digital library access fees. Pricing was set to encourage electronic access.
Member pricing was kept as low as possible, using a cost-plus calibration rather than a what-the-market-will-bear approach.
Both membership dues and subscription revenues were at risk through the cancellation of print and through the potential substitution of institutional licenses for individual subscriptions. There were unknown costs for building and maintaining an electronic resource: post-print production processes; online customer service; new cost-of-sales for institutional and consortia licenses; functionality enhancements; and infrastructural evolution. Some were duly anticipated; others not; all were exposures that required close analysis.
Careful and conservative articulation of a business model resulted in a financially viable digital library from its very start—a very modest but nonetheless startling success in light of other similar online endeavors that enjoyed the luxury of far greater up-front investment but suffered from much slower market uptake.
Today, vigilance is required to guard against runaway costs for enhanced functionality and expanding content; a constant battle between the allure of what can be done and the pedestrian pragmatism of what makes economic sense.
The need for a thorough-going system rewrite often coincides with the advent of enterprise computing. Disparate resources with incompatible representations of overlapping and redundant data are perfectly rational in a world of printed products supported by independent, heterogeneous applications. The coordination of these separate products is effected by intelligent human communication. But when all this information must be integrated in a coherent set of programmatically generated Web pages, such as on a corporate site with a branded identity, then disparities among the underlying information resources result in what might be called a "Talmudic" programming effort. The resulting system then requires all available resources simply to keep it running. Nothing and no one is available to enhance the system and rationally plan its future evolution. When such a system is developed hurriedly, as most are, its maintenance costs are multiplied by what Robert Glass referred to as "fragile code"(See www.acm.org/pubs/citations/journals/cacm/1998-41-12/p11-glass).
These are the typical conditions that demand a thorough-going system rewrite, where coordination is ensured by integration at the level of the data model.
Join the Discussion (0)
Become a Member or Sign In to Post a Comment