Logo IMG
HOME > PAST ISSUE > Article Detail


Avoiding a Digital Dark Age

Data longevity depends on both the storage medium and the ability to decipher the information

Kurt D. Bollacker

… Or Fake It!

Once we’ve thrown the dice on our data-representation choices, is there anything else we can do? We can hope we will not be stuck for decades, like our NASA archivist, or left with a perfectly readable but incomprehensible Phaistos disk. But what if our scattershot strategy of data representation fails, and we can’t read or understand our data with modern hardware and software? A very common approach is to fake it!

If we have old digital media for which no compatible hardware still exists, modern devices sometimes can be substituted. For example, cheap and ubiquitous optical scanners have been commonly used to read old 80-column IBM punchcards. This output solves half of the problem, leaving us with the task of finding hardware to run the software and interpret the data that we are again able to read.

In the late 1950s IBM introduced the IBM 709 computer as a replacement for the older model IBM 704. The many technical improvements in the 709 made it unable to directly run software written for the 704. Because customers did not want either to lose their investment in the old software or to forgo new technological advances, IBM sold what they called an emulator module for the 709, which allowed it to pretend to be a 704 for the purposes of running the old software. Emulation is now a common technique used to run old software on new hardware. It does, however, have a problem of recursion—what happens when there is no longer compatible hardware to run the emulator itself? Emulators can by layered like Matryoshka dolls, one running inside another running inside another.

comments powered by Disqus


Subscribe to American Scientist