Standards are the Sands of Time (or, the Illusion of Preservation)

On a long enough timeline, the survival rate for every standard drops to zero.

From emulated games to digital negatives, how can we preserve our digital content for posterity, and inform our decisions about which formats we adopt?

Commodore VIC-20 and Tape (deceased)
A special jpg: 2 monkeys, city, moonlight
A special jpg: 2 monkeys, city, moonlight

How many jpg images do you suppose are living on the internet? Probably quite a few.

Among all those images, the jpg on the right is particularly special to me, because it’s a screen capture of an adventure game I wrote for the Commodore VIC-20 when I was very very young (like almost single digits).

If you can’t make it out, those are two monkeys standing under the Toronto skyline, reading a newspaper by moonlight.

At the time I wrote the game, it could only be played by reading the bits that were encoded as waveforms on an audio cassette tape. Which was awesome.

But one day my VIC-20’s cassette drive silently stopped reading tapes, and the game could be played no more.

The photograph below (another jpg) was taken of the defunct
VIC-20 hardware some time before it was tossed into a big dumpster.

Commodore VIC-20 and Tape (deceased)
Commodore VIC-20 (deceased) and Cassette

Resuscitation by Emulator

VICE Vic-20 Emulator on Windows
2001 Emulator (not dead yet)

Fast-forward through more than a decade to 2001, where I find VICE, a program that can emulate the VIC-20 on Windows.

With the game’s cassette reclaimed from the basement and a dusty tape player connected to my PC’s audio port, I transferred the audio that made up the game’s bits into an audio .wav file, and then converted that audio file into a format that could be loaded by the emulator.

I was so happy to be able to run my game again, and have it preserved “forever,” even though its contents are meaningful only to me (and maybe my family).

Satisfied, I zipped it up together with the emulator
and put it into my folder called “C:\pastlives” for safe keeping.

Now almost ten more years have passed. In this 64-bit world, my game still runs.

For now.

(slightly) Longer Term Thinking

This ran through my head when I was considering whether I should adopt Adobe’s Digital Negative (DNG) format as a standard for my photo archives. The alternative is the Nikon raw (NEF) format I currently use for my negatives.

It literally kept me up last night thinking about the layers of technology I depend on to retrieve and view things in “C:\pastlives”, like my VIC-20 emulator or my digital photos.

Check out this stack of just some of the standards and technologies I used to play my game today:

Some of the technology standards used to play my VIC-20 game on Windows 7
Some of the technology standards and file types used to play my VIC-20 game on Windows 7. (another jpg)

(Red == the ones I perceive as more volatile)
Sure, you could substitute for some of these (the VIC-20 emulator, for example, also runs on Linux).

But how likely is it that a complete cocktail of prerequisite technologies will be around long enough for me to load and play the game in three more decades? In ten?

These standards aren’t stacking up like a solid foundation, they’re piling up like sediment!

(To abuse a pretentious analogy) It’s like the sands of time themselves.

Consider the Floppy

5 1/4 inch Floppy Disk
5 1/4 inch Floppy Disk

Remember floppy disks? Those once-ubiquitous storage devices? How many of you still have machines around that can read a 5 ¼ inch floppy?

On a long enough timeline, the survival rate for every technology standard drops to zero.

By my estimation, all it’ll take is for me to neglect its preservation for approximately a decade, and my game shall be played no more.

Spoiler Alert!

I’ll go out on a limb and say that despite my efforts, my VIC-20 game will never outlive even all the jpgs in the internet.

So I’m just going to tell you that one monkey rescues the other monkey from the Toronto zoo and they escape back to Africa.

There’s an action sequence where you have to climb through the sunroof of a taxi and jump off at just the right moment to escape a high-speed police chase.

Screenshots, while they last, are available upon request.

5 thoughts on “Standards are the Sands of Time (or, the Illusion of Preservation)”

  1. Instead of attempting to preserve the reader/interpreter for the code why not port the code/data? If you can make a bridge from Tech1 to Tech2 and then Tech2 to Tech3 and so on you could form a hopefully automated and resilient history. Though some things may be lost in translation, the hues of your blues and the subtle edges of your monkies. I’ve got some music hanging around that has gone from cassette tape to CD and now to MP3. There are subtle differences but nothing I can honestly make out. Over a 1000 re-encodings though I imagine the differences would become apparent.

    Have you seen the Long Now Foundation? 10,000 year plan.

  2. I like the idea of porting content between technologies in theory, but in practice I think the effort only gets invested for content like The Secret of Monkey Island but not my lowly monkeys.

    Your cassette-to-CD-to-MP3 example is great and reminds me that in “C:\pastlives” I also have some audio stories my sisters and I recorded as kids that made the same progression. I love the fact that the low-fi cassette recorder hiss has been preserved. The MP3 hop made me kind of sad, because it was lossy and not in a way that I thought added texture. But you’re right, I am just grateful that the content is still around, even with some of the proverbial edges subtly updated.

    I read about the Long Now Foundation in relation to a Wired article on Danny Hillis from some years ago. I’d love to attend one of their ‘seminars on long-term thinking’.

    10,000 years pass … my monkeys outlast humanity … wasn’t there some Charlton Heston movie like that…?

  3. I welcome our new 8bit monkey overlords!

    The iPhone Monkey Island example made me think a bit. I was thinking more generalised porting, something automated that can be run against Language X to produce Language Y. That already exists though and the results are rarely good enough. Time and effort and attention to the new destination’s capabilities and limitations (iPhone) is required. Audio has a fairly wide acceptance range when transcoding but computer code doesn’t. Hmm. Tough one to tackle.

  4. I am still thinking about this, and am not sure my analogy to sediment is quite what I had wanted to articulate.

    What I am wrestling with is that the ‘foundation’ itself is destined to become the obscurer. So I wasn’t meaning to be pithy with the ‘sands of time’ analogy.

Comments are closed.