! Copyedits from BW, revisions per review from WA, corrections, AF 6/8/98 3:37 PM >
From the Editor
Running Code and Paths Not Taken
Four stories in this issue are about implementation projects: Timothy Thomas's story on the technical basis for the Physics Review Online Archives (PROLA); Michael Roszkowski's story on a distributed architecture for metadata under the aegis of the Scout Project; Richard Goerwitz's story on pass-through proxying as a strategy for managing secure off-site access to support remote learning at Brown University; and Andy Powell's story on using Squid as a way of resolving Universal Resource Names (URNs) at the UK Office of Libraries and Networking (UKOLN). All of these systems are either up and running (PROLA, Brown, and Squid) or are proposals to establish working, albeit experimental, systems based on existing projects (Scout). In a fifth story, David Bearman and Jennifer Trant discuss requirements for museum systems, making the implicit assumption that such systems are being built -- as, in fact, they are. The sixth story by Seamus Ross and Maria Economou on the role of information and communications technologies in cultural heritage programs asks the questions, when should these systems be built and what is their scope?
The value placed on working systems embodied in these stories has been inherited by digital libraries research from computer science, where, for example, the acid test among the working groups of the Internet Engineering Task Force (IETF) has been running code. Arguably, computer science inherited this value from the culture of engineering as distinct from its parentage among the more theoretical disciplines of mathematics and physics. It resonates, moreover, with the culture of librarians who answer questions today and need systems that will work tomorrow.
The underlying preference for the practical over as the theoretical places priority on evidence in the form of experimental systems and testbeds, and means that progress is frequently incremental, building off existing and legacy work, whether or not this is intentional. We all use QWERTY keyboards, a well-known example of late nineteenth-century technological lock-in, because they exist and are ubiquitous, not because human factors research has demonstrated the value of the configuration of keys. And software designers worry -- rightly -- about forward migration as well as backward compatibility. Why? Partly to address potential commercial viability, partly to reduce or eliminate redundant effort, but also to ensure that the code will run. Indeed, these themes form a leitmotiv through many of the stories in the June 1998 issue of IEEE Computer Society's Computer, devoted to using commercial, off-the-shelf software (COTS).
The attention to running code leads naturally to a sensitivity to legacy systems and an awareness of the installed base. It means that progress in research tends to be incremental rather than obviously revolutionary. Given an array of choices, the unspoken bias will be to work with existing computer languages, for example, rather than continually inventing new ones, unless the existing choices are obviously unworkable. Of themselves, standards, whether formally adopted or arrived at by consensus, do not mean insufferable constraints: A lot of hardware runs on the 110 volt standard even though the 110-volt decision in electrical systems arose from an analysis of the economics of copper and construction in the United States in 1880. And even our units of analysis -- volt, ampere, and hertz, named for the physicists who worked out the relationships: Alessandro Volta (1745-1827), André-Marie Ampère (1775-1836), and Heinrich Hertz (1857-1894) -- embody basic research of the 18th and 19th centuries.
Still, choices taken mean avenues left unexplored. Sometimes revisiting assumptions and choices can lead in productive directions. Reginald Ferguson's re-reading of James Clerk Maxwell's 1873 Treatise on Electricity and Magnetism led to his innovative continuous wave radio technologies, developed between ca. 1900 and 1906, which represented a departure from the spark technologies developed by Oliver Lodge and Guglielmo Marconi in the 1890s. On the other hand, my computer, the lights, and my telephone work just fine in a 110 volt world. Sometimes, Robert Frost tells us, two roads diverge, "worn really about the same." Whichever we chose, we will -- and should -- always wonder where the other led.
Copyright (c) 1998 Corporation for National Research Initiatives
Search | Author Index | Title Index | Monthly Issues
Previous | Next
Comments | E-mail the Editor