I think there's room for substantial innovation in the face that libraries present to the world," says Tim Bray, the architect of a program that can turn a library's Web catalog into a browsable, customizable map. The program supplants the search-and-retrieval model of the online public access catalog with a graphical map that can show how a library's varied resources--print books, e-books, video, audio, and Web databases--relate across subjects and categories.
"Since the first person had the idea of building an OPAC, all OPACs more or less look like other OPACs. They get faster, and the search gets a little smarter, but they're basically the same thing. You have to wonder, is it the case that we got it right the first time that we tried it? That seems unlikely," says Bray.
Bray's technology resume is long and impressive. He is most widely known as one of the primary developers of XML, a text mark-up code for Web and other media applications, and the architect of OpenText, one of the first Internet search engines. Bray is also the architect of VisualNet for Libraries, a program produced by Vancouver-based Antarcti.ca systems. It is the first product in what may be the next wave of Web-based applications, software designed to expose the organization of the library catalog and make steps toward a browsable OPAC.
Enthusiasts of knowledge-visualization software believe that current OPAC interfaces, which do not allow for significant browsing, have driven student researchers to Web-only sources. By encouraging browsing within the levels of a Web-based data map, products like VisualNet may reveal the richness of library collections and entice students and patrons back into the stacks.
Can one graphical interface accomplish all that? Mapping applications appear to have a rich future in libraries, but in the beginning, libraries weren't even on the VisualNet radar.
Corporate concept, community application
Bray turned his attention to the problems of interfaces in the late 1990s. The arrival of the Web browser in the mid-1990s transformed the way that people used the Web, but, Bray observed, Web interfaces still had not caught up with the graphical model of the traditional Windows or Mac desktop.
"There's a yawning gulf between what you have for your desktop and what you have for your shared information. You have folders and trash cans, and you have visual access to everything on your hard drive. As soon as you get off your hard drive, you're essentially in the mode of typing queries and getting lists of results. There is so much out there, and I wondered, why is it so nonvisual?" asks Bray.
With a plan to develop and market data maps for corporate intranets, Bray founded Antarcti.ca systems in April 2000. (Before that, says the company's communications director, Stephanie Lummis, it was "a little skunkworks in Tim's basement.") Map.net, a navigable map of the Internet, was launched in November 2000 as a proof of concept and a marketing tool. When the map, which was shaped like Antarctica, garnered attention in the press, the company started hearing from the library community.
"A couple of librarians called us up out of the blue when we got some coverage for our product launch, and said hey, we could use this," explains Bray.
The librarians' hunch made Antarcti.ca look closer, and caused the company to create another model, plotting part of the PubMed database for the National Library of Medicine. In the enterprise sector, Antarcti.ca had discovered that not all data makes a good-looking map. Many companies had disorganized data that, when run through VisualNet, made an unattractive diagram. Libraries, however, have massive amounts of hierarchical data that can be easily formed into a visual model. Says Bray, "I have to be honest, we weren't smart enough to think of it ourselves."
Bray, who cofounded OpenText in 1989, has long kept an eye on libraries. …