The growing number of documents being preserved electronically is a pressing concern for archivists.
“Digitization is great for moving data around, but it’s not yet known what happens to data when it is stored for long periods of time,” says the University’s chief archivist, Michel Prévost.
“There are two matters that are especially worrying. One of these is concerned with assuring ourselves that we will be able to read the documents we electronically produce in 10, 20, 30 or 50 years. And secondly, we need to ensure these documents are properly named.”
For example, when the University migrated to Microsoft Office, very few people took the time to resave their old WordPerfect files in the new MSWord format.
Prévost emphasizes that it is practically impossible to read a file produced with WordPerfect 4 with today’s software without losing a large portion of the native information. Unfortunately, he says, whole chunks of information are already irreparably lost to the University because they are stored on obsolete data-carriers that can no longer be read.
“For these reasons, we need to automatically migrate data— like we do with the central servers—whenever software or equipment is upgraded,” insists Prévost.
Moreover, the Archives department often receives electronic documents, on diskette for example, which are not clearly identified. Sometimes, the file names are very generic, such as “Project1, Project2, Project 3,” and as such, do not tell very much about the files’ contents. In addition, passwords known only by ex-employees of the University are commonly encountered, making it impossible to open certain documents.
“Many important decisions at the University are currently captured by email,” says Prévost. We typically see one of two things here: people who don’t save any email or people who save everything.”
In the first case, all traces are erased; in the other, a sea of data remains, which becomes almost impossible to comb through in order to extract the information worth preserving.
Previously, recalls Prévost, it was possible to copy a microfilm document and destroy the original, thus reducing the space necessary for storing files. “With digitization, this is not yet possible,” he says, since the quality and the integrity of the preserved document is not guaranteed.
In addition, data corruption sometimes occurs quite rapidly. Diskettes, in particular, can be easily affected by magnetic fields, heat or moisture. Even CDs slowly degrade over time.
Web pages present a particularly unique problem. As the Web is in a perpetual state of flux, web page updates are constantly performed without preserving any copies of previous versions. When navigating the maze of cyberspace, we thus find ourselves locked into a perpetual present, and essentially disconnected from the past.
Related Links:Records management Archives of the University of Ottawa
