Ephemera Are Us (Newsletter)

This is a story about why technological innovation should worry us as our lives inexorably shift online, if only because innovation begets obsolescence.

Fourteen years ago I launched my first website. In that pre-milennial, pre-Google epoch it was standard fare, just a bunch of HTML files with links and images, some animated. I had just acquired a doctorate in Geography, and this platform would serve as my online geospatial consulting shingle. Its content spanned some 50 Web pages, including several hundred images and about a dozen downloadable documents. Not bad for a debut website in 1999.

The curious can visit it here. Click anywhere to enter. It’s geo-geek stuff, but that was me back when. See the highlight reel, an animated comic strip summarizing my doctoral research (plays best on the Chrome browser).

So that’s my little geospatial museum. As far as I know, the site was never hacked, although the web server it shared space on may have been. That’s probably because my site had no database or computer code to subvert.

You might notice the pages have geometric mosaics on their left sides, unfortunately repeated towards the right. That was a bad design decision, based on my old computer’s screen resolution. I can’t easily fix it, as the graphics for those patterns are stored in files from a program (called Canvas) that no longer runs on today’s operating systems. And that illustrates the point of all this: All that groovy stuff we expect to find on the Web depends on hardware and software that is forever being updated to higher levels of complexity that have new ways to be hacked or otherwise fail.

Here is a more recent and less fortunate example: I was an avid user of a storytelling site called cowbird.com that collected over 88,000 stories from 40,000 users between 2012 and 2017. This is what it looked like on one particular day:

Cowbird was enchanting, easy to use, and very sociable. Many friendships formed among the writers who posted, loved, commented, and collected each other’s stories, categorized and tagged to make them easy to look up and associate. Most had images and some had audio tracks.

Today cowbird.com only exists as digital flotsam in the Internet Archive. It became a rudderless ship in 2017, when Cowbird’s creator, a polymath media artist named Jonathan Harris, wanting to go on to other things, decided to stop accepting stories. At the time, Jonathan said he intended to keep the site open as “a library of human experience,” but soon after entering read-only mode it tended to go down. Whenever I noticed it was offline, I alerted Jonathan and asked him to kick-start it. The last outage was in February. After it had been dark for a month, I informed Jonathan, who responded:

The hosting company that had been donating our servers since 2011 … discontinued their pro bono support for the project. Also, the version of PHP that we used to build Cowbird is now 7+ years obsolete, and we’re pretty sure our servers were recently hacked and malicious code inserted that keeps causing the servers to be automatically taken offline by the hosting company as a precaution. .. my sense is that there may be no simple way to get it working again in a sustainable way. After seven years of keeping the archive online, it may not be practically possible to do so any longer.

Cowbird’s fate is far from unusual. Information on the Internet may seem eternal, but as true as that may be, the software and systems that access its data are anything but. There are websites I cannot read because my browser is too old and can’t be upgraded unless I buy a newer computer. (See more of my obsolescence woes here.) Content I bookmarked over the years has gone 404 (i.e. “The page you are looking for can’t be found”). Some sites I’ve revisited are up for sale, like those empty shops you see in malls.

Back in 2013, I worried about the ephemeral natures of digital media in one my Cowbird stories. To make that point, my story used itself as an example of how the written word has become more code than text, code that we depend upon machines to interpret for us. Here’s a screen shot from the story of what its HTML code looked like. What I had written (that grey band) comprised less than half a percent of its content.

I had to dig through the Internet Archive,^* aka the Wayback Machine, to find a snapshot of that story, and then its audio track wouldn’t play. As it’s one of my favorites, I updated my original manuscript and managed to persuade the editor of bookscover2cover.com to rebirth it, retitled Electrifying Oral Tradition.

Its premise is that storytelling in pre-literate societies was much more like electronic than print media. Remembering stories, recalling them, and passing them on required only human speech and neurons (self-programming software), without recourse to hard copy. Of course, people can forget stories or garble them, as also happens on the Web. But books don’t, and are likely to last a hell of a lot longer than digital media. (Although they, too, can get ruined or destroyed; remember Fahrenheit 451?)

Before submitting my story, I updated it with a transcript of session I recently had with the celebrity AI conversationalist, ChatGPT, whom I asked to tell me a story. Try that yourself: sign in and simply type “Tell me a story in 100 words or less.” My prompt to ChatGPT was more specific, and what I got back was a little unnerving, as you’ll see if you read the essay.

Before posting stories to cowbird.com, concerned that it might not outlast me, I drafted my stories in Microsoft Word, ending up five years later with 13 files containing 450 stories spanning 1400 pages. I considered printing them all out, just in case something bad happens to my computer, but rather than sacrifice a tree, I’ve backed them up and will take my chances with the forces of destruction.

If your digital devices hold important memories, please find a way to safeguard them for future generations.