Subscribe
Subscribe
MY AMERICAN SCIENTIST
LOG IN! REGISTER!
SEARCH
 
Logo IMG
HOME > PAST ISSUE > Article Detail

COMPUTING SCIENCE

Bit Rot

Brian Hayes

The $1.29 Oeuvre

Are all those bits and bytes worth saving? Surely not. But if they're not worth saving, they're also not worth throwing out. The cost of magnetic disk storage is roughly 10 cents per megabyte these days. Tapes and CD-ROMs are even cheaper. My entire oeuvre—everything I've ever written for publication, as well as all the private and personal ephemera of a lifetime, from school compositions to love letters to grocery lists—all this would fit on a single CD-ROM. The storage cost for a lifetime's worth of words is $1.29. That's not much incentive for cleaning out the attic.

Lately I've found an even better reason for keeping those files. I've discovered that I'm not just a writer anymore—I'm a content provider! And my disk drive is stuffed full of content.

In years past an author's final product was the printed page. The computer file was nothing more than an intermediate stage in the process of putting ink on paper. Once the presses were rolling, the disk files had no further value, and they could be discarded just as carbon copies were in the age of the typewriter.

All that has changed. Print is no longer the only destiny of the written word. This column, for example, will not only be bound into a printed magazine but will also be posted on the American Scientist Web site. Indeed, it will be offered in several electronic forms, including Postscript and PDF files that mimic the appearance of the magazine pages and a version coded in HTML, the native language of the Web. Someday the same column might be made available in still other media, such as CD-ROM. An abstract might be prepared automatically for a bibliographic database. The article might be reprinted in an anthology with a different typographic design, or it could be reproduced through a print-on-demand service for classroom use. Who knows what else might be in prospect? Last year's printout and this year's Web page could be next year's virtual-reality environment. Or tee-shirt.

In this new world of digital content, computer files are most certainly not disposable intermediate forms. The disk version is the master document, from which everything else derives. I might well throw away printouts and proofs, but I keep the disks. Furthermore, I worry about files that might be stranded or orphaned as computer hardware and software evolve.

This is not just my problem (although I may be more compulsive about it than most). In the sciences, almost everyone is becoming a content provider. Papers are submitted to journals and conferences in electronic form, and they may also appear as electronically distributed e-prints. Supporting data, such as genetic sequences, wind up in public databases. Many of these digital documents are considered a permanent part of the scientific literature. Their life expectancy is greater than that of the software that created them.








comments powered by Disqus
 

EMAIL TO A FRIEND :

Of Possible Interest

Feature Article: The Statistical Crisis in Science

Computing Science: Clarity in Climate Modeling

Engineering: Anonymous Design

Subscribe to American Scientist