The Non-Durability of the Web: Yahoo! discontinues GeoCities

At the end of October 2009, GeoCities will cease to exist when Yahoo! pulls the plug on the service that helped wanted many personal websites back in the 1990s. All is not lost however: the Internet Archive is undertaking an effort to store as much of this material as they can, but it is unclear how they are doing that or if they should. In many ways, GeoCities is past its prime but there is still plenty of interesting content there and it shows how the early days of the Web operated. Comparing Web usage from 1997 to 2007 could make for an interesting research project, I imagine.

The Internet Archive’s efforts to ‘archive’ GeoCities, however impressive, cannot be considered “archiving” in the professional or classic sense. The relatively low cost of data storage (low is not the same as zero or free though) seduces some to think that simply everything produced should be stored – it would appear that is how the IA is ‘archiving’ GeoCities content. Simply making copies of as much content as possible is not what I would call an archive. Imagine an organization is moving out of a building and setting up elsewhere, would it make sense to “archive” every single piece of paper and data in the office? Such an archive would be both large and difficult to use. What criteria are being used to determine what should be kept? What about the people who created the GeoCities content?

The article linked to above raises some interesting questions about the durability of data on the Internet. Institutions such as universities and libraries have, in many cases, existed for decades or centuries. It is unlikely that a dissertation or other valuable item at Harvard or Oxford will be in danger of loss (though fire, flooding and other disasters are always a possibility), but is that the case for free Web services? If the example of the sudden death of GeoCities is any indication, then one has some reason to be skeptical about the longevity of data in cloud applications such as Gmail. Maybe there will be a move, at some point, to introduce “data longevity” standards into user agreements? Maybe such guarantees will serve to differentiate free services from paid ones? In the final analysis, it may only be traditional archives that can be counted on to archive content professionally and retain it for the long term.

No related posts.