Issue No. 01 - Jan. (2013 vol. 46)
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/MC.2013.32
David Alan Grier , George Washington University
During a recent spate of unpleasant weather in New York City, I contacted my friends Todd and Sharon to see how they were surviving the difficulties. Sharon quickly replied that neither time nor space was interfering with their operations.
"We pulled out of the city the day before the storm," she replied, "We're at our place upstate and are just as productive as if we were in Manhattan."
The storm kept them away from the city for at least a week, as it damaged the power grid in their neighborhood and flooded the ground floors of many buildings around their offices. It also emphasized how much the Internet has helped us conquer the limitations of time and space.
Barely two decades ago, companies were tied to email and servers located in their own offices. A firm couldn't abandon these servers to a hurricane without abandoning its work as well. Under such circumstances, a week in upstate New York wouldn't be a productive period of work but a vacation marred by inclement weather.
The same technology that has caged some of the power of geography has also limited the ability of time to wither and narrow our lives. Thinking of my friend Sharon's situation led me to a quick Internet search that revealed more than a few remnants of her prior career as a punk rock musician: videos, lyrics, gig announcements, photos combining the high fashion of 90s punk with its implied threat of physical violence. If you didn't look at the dates carefully, you might conclude that her band was still performing regularly at clubs that vanished from the city long ago.
The persistent memory of Sharon's musical career comes from our efforts to make the Internet a permanent repository for the world's knowledge. Almost from the start, the pioneers of the Internet argued that all network information should be permanent. "Pretty much the only good reason for a document to disappear from the Web," claimed Tim Berners-Lee, "is that the company which owned the domain name went out of business or can no longer afford to keep the server running."
Of course, documents disappear from the network for many good reasons—they might be no longer relevant or describe activities that have been overtaken by events. They could be preliminary information that has been replaced by a final version. But even if we didn't encounter such challenges, we would still have tremendous problems keeping the network as a permanent repository of information. Software changes. Organizations do go out of business. Ideas are reorganized into a more convenient form. "As is the case with handling change in general," noted some Computer Society members in a recent issue of Internet Computing, "managing change in a linked data environment turns out to be hard."
Equal to the problem of managing change in a linked data environment is the task of interpreting the information it generates. Documents carry dates that mark when they were created, but these dates might not reveal much about the context of the information. All we know is that they fall after the date for the information in the document. Sometimes it's easy to put the documents in context: a picture of Sharon as a baby is certainly older than those of her as a business executive, but those of her as a punk rocker might inadvertently seem contemporary to those she recently posted for her company.
With computer systems offering the Internet's organized knowledge to our applications, we no longer live in an age where past is prologue. The past is now intermingled with the present and is sometimes indistinguishable from it.
David Alan Grier is the author of many books, including the recent The Company We Keep (IEEE CS, 2012). Contact him at email@example.com.