Digital immortality is a myth, what can we do about it?
I recently watched the finale for The Amazing Digital Circus. I won’t spoil anything in this post, for those who haven’t had time to watch it yet, nor will I share my opinions in depth.
However, watching this finale, it got me thinking about the ideas of bitrot and internet content degradation. It is a fascinating concept to me. One day, you will die and be forgotten. The only thing left of you will be whatever footprints you left, and I think most of us have left plenty of footprints on the internet. How do you store data in such a way that it will be accessible for future generations?
This is quite a layered problem. If you post something online, what’s in the way of your post staying up? The first layer, I’d think, is whatever web hosting you’re on deciding to stay up. Web hosts go down, all the time. Anyone who’s ever browsed a forum can attest to the legions of dead MediaFire and PhotoBucket links. Some part of the history of SCP is lost after the EditThis wiki went down. Just a week ago, it was announced that the RPGMaker forums were going down to be replaced by a new “guild” forum.
There’s quite a few reasons why a webhost might go down. Could be running out of money, or they got tired of hosting it, or plain old corporate greed.
Yes, I know, the venerable Internet Archive exists. That being said, the Archive is intensely fragile. All it takes is a bad year or a lawsuit gone awry for the Archive to be lost forever.
The second layer is incompatible formats. I recently encountered this problem for myself. I was helping my mom clean out the garage, and we found a bunch of floppy disks that she used to store assignments in college. Of course, none of our computers have a floppy disk reader. What was once the bastion of long-term file storage is now effectively dead weight.
This also extends to file formats. Flash ruled the world not even 20 years ago, but is now a decadent fossil. Even Ruffle fails for quite a few old Flash files. Entire libraries are now inaccessible because of format drift.
Regardless, this brings us to the third layer: hardware itself. Hard drives and SSD’s alike have limited shelf life. Anything except for a top-of-the-line hard drive will last no longer than 20 years before big sectors start becoming unreliable. A hard drive kept in storage needs to be continuously replaced every now and then. This requires procedures, schedules and institutions. All of which are known to crumble.
I don’t have a source for this, but I’ve heard a theory that this period of history will become a “dark age” as digital data crumbles away. The internet of old will be lost, left as an incomplete set of unreadable web formats on crumbling hard disks. Future archeologists will be left to reassemble what our culture looked like from whatever survived as a physical copy.
I’m somewhat more optimistic than that, but I do believe that ~80% of the internet will eventually become lost media. There’s only so much that can survive.
As for me, I used to host this blog behind an nginx server, before I realized it was all static and shoved it onto [Cloudflare Pages]1. This decision was largely motivated by wanting a page host that would require less of my continuous attention. If push comes to shove, I’m happy to go back to nginx, even if it’s somewhere local. At some point, I might print out my blog, bind it all in a suitcase, and bury it in my backyard.
Data will rot. That’s okay, that’s part of accepting death. All we can do is all we can to ensure it becomes the next generation’s problem.
-
This is not an ad. ↩