One day, I noticed that a couple of video files we created a few years back wouldn’t play. At first I shrugged. Gradually, more and more files stopped working, on a variety of Web browsers.
Every few centuries, great collections of human knowledge go up in flames or artistic masterpieces go down before the knives and hammers of fanatics and barbarians.
In late antiquity, an irreplaceable collection of ancient writings in Alexandria Egypt’s famous library vanished in one or more acts of wanton destruction. In medieval times, the Mongols took days to destroy the collection in the Grand Library of Baghdad. On this side of the ocean, exactly four Mayan books survived the bonfires lit by Catholic priests. Today the devastation continues in Syria and Iraq as the savages who call themselves ISIS cleanse what they consider to be profane art and learning.
For our little video project no one would care, in a century, if it were gone, save a historian of 21st-century universities — or perhaps fashion. I was also sure there would be a workaround to fix our files. Yet I began to think about the petabytes of digital content that might matter, once our era’s creations no longer functioned with the tools our descendants might use.
We’d created a few hours of videos, all in short clips of a minute or two, featuring professors talking about the value of writing in their academic fields. Many of them made points so strong that I’ve shared them repeatedly with my own students. We thought our content equally strong; we used industry-standard video coding and compression, in a common format. We hosted everything not on the cloud but on our own server in a bombproof data center. We kept multiple backups in several places. As I later discovered, none of that may matter when a company updates its software. With one operating-system upgrade, our clips gradually stopped playing on the Web, in both Windows and Mac-OS computers. There was no rhyme or reason to this ghost in our machines. Some browsers ran the videos fine, while on a different computer the same browser would not, even with the proper “plug in” installed.
Our creations were going dark, one little segment at a time. Every backup had exactly the same problem and the original raw footage had partly gone missing. Many hundreds of hours filming, editing and coding the pages were in jeopardy. The software provider’s own solution did not work. Meanwhile, one of our back copies had been destroyed by a hardware failure.
The story finally had a happy ending, when I talked to our campus technologists about the most durable cross-platform video “codec” to give our content longevity, at least for a while. I can code HTML the old way, so I dove in like a monk repairing a Latin manuscript he knew by heart, after it had been damaged by a rodent. I began changing file names and removing obsolete “tags” deep in the bowels of the code. In our tests, videos began to play correctly again. We have a few dozen hours of coding ahead of us still. The fix worked.
But would it next time?
I imagined myself in a decade, called back to work by my successor, to recode some of the videos yet again. I’m not sure I could. The one computer capable of fixing the current problem is now a decade old, a machine so ancient under Moore’s Law of computing progress that it’s no longer secure enough to be networked. I kept it because it had a once-wonderful piece of software dumbed-down in the maker’s subsequent “upgrades.” So I found myself manually transferring files to an external hard disk and walking to the old trusty machine to load the videos for conversion. In academia we call this a “Sneaker Net” connection.
Our problems were small when compared with the now-forgotten Y2K crisis, when we narrowly avoided a tremendous disaster globally. Banks, insurance companies, defense contractors and government agencies began hauling gray-hairs and bald-heads out of retirement to fix errors in old Cobol and Fortran code. I knew one such coder, now deceased, who told me about the gravity of the problem he faced when he walked back into his old office. Had the Y2K bug not been addressed in time, he predicted a collapse of government payrolls, taxation databases, banking records and civilian infrastructure lasting “a long, long time.”
Readers here may damn one software provider or another, but in my experience the big software houses such as Microsoft, Apple, Google and Adobe have all made changes that endanger knowledge we’d like to leave for our descendants. These are the ephemera that may go up in smoke like books in Alexandria. Our data on the cloud, from family photos to poetry shared in blogs, may vanish in a cyber attack.
For those who might go back to paper, today we aren’t simply talking about text and images, which, in earlier times, could be preserved in that durable medium. One cannot print a digital video or an old piece of gaming software. Perhaps the best way to address this is through a consortium of industry and government, like one being built by the wise people at archive.org.
We must be certain we all have digital formats that can endure long enough for content creators and archivists to make regular updates, as well as legacy hardware and software to understand our dusty old creations. We also need the equivalent of the Cold War’s Mount Weather or today’s seed banks: an apocalypse-resistant, international archive in a secure space to preserve vital information about our arts, culture, science and technology.
I do not fear a sudden collapse into barbarism and burning libraries. But I do fear that we may lose a digital equivalent of Van Gogh’s “Starry Night” or Wendell Berry’s poetry. And those would be losses for the ages. S
Joe Essid teaches writing and fumbles with code at the University of Richmond.
Opinions expressed on the Back Page are those of the writer and not necessarily those of Style Weekly.