While the rest of the Web-savvy world fawns over breakthroughs in real-time search and pontificates on the future of social networking, Los Alamos National Labs is looking to the past. A team there is developing a “time traveling” Web browsing technology, dubbed Memento, that will allow users to find old versions of Web pages without trolling old archives.
Current archival resources, like the Wayback Machine, can search for old, cached pages of Web sites, but a search for a particular URL brings up little more than a list of cached dates the user must sort through to (maybe) find what he or she is looking for. Like an online gull-winged Delorean, Memento will take you directly to the web page you want at the historical time and date you specify, by harnessing the content negotiation feature of Hypertext Transfer Protocol — HTTP — the system that serves as the infrastructure for the entire Web.
One of HTTP’s central functions is content negotiation, during which a URL can send various types of data to a browser depending on what the browser asks for (this is why a Web server will often send a Spanish version of a site to someone in Spain and an English version to someone in the UK). Memento simply adds a new query layer to the its request from a server, asking it to call up a specific time and date as well.
Using a server with built-in Memento — so far the Los Alamos team has developed a plug-in for Firefox as well as a version with Memento built-in — users simply select a date and time from a drop-down box and go to the URL as they normally would. The browser goes straight to the archived page for the date and time specified, no archive wandering required. Naturally, for Memento to be most effective, site-owners will need to keep more detailed archives with more time-stamped pages. But the Los Alamos team thinks that a better way to find archived information will encourage site managers to archive more pages.
Honestly, we just want to have one last dance with GeoCities.