"A lot of the current examples don't really work. They're glorious failures�just eye candy. You look at them and say, "Oh yeah, that's quite interesting, but I wouldn't ever want to really use it." Despite all those fancy maps and experimental 3-D stuff, I'm still surfing the Web using Netscape. I think and hope it's just a matter of time before some smart grad student at MIT or Stanford comes up with a killer map of cyberspace that changes the way people view the Web. One niche for innovation is in the design of search engines. The typical search engine still looks pretty much like it did five years ago, when Lycos and WebCrawler first started. You have a box, you type a bunch of keywords, and you get a long list of items back. I'm constantly amazed that somebody hasn't come up with something better. It's a lack of imagination. But I think it can be done."
From Charting The Virtual World, in which Darwinmag talks to Martin Dodge, co-author of 'The Atlas of Cyberspace'. Some nice images from the sidebar link, but I'm not convinced about Dodge's desire to see some killer-app mapping of search engine interfaces or email clients. Sure, spatialisation etc. can throw extra light on navigation, but perhaps there's an equal and opposite case for ever more focused renditions of todays text-based search/organisation interfaces. Sometimes our desire to map may just be the desire to make the eye-candy he mentioned earlier. They're not mutually exclusive solutions of course, but I keep thinking of dreadful VRML 'worlds of data', and comparing that to how efficiently the average punter now parses a list of search results. Would those simple text relationships be improved by spatial/visual metaphors?
