Keeping stuff: How to preserve course papers despite technological change

Disk storage

In 1982, I took some more courses in computer science, and for the first time I had the opportunity to create and save documents and programs electronically, in a medium not based on paper. It didn't amount to much; the share of hard-disk space allocated to each student in an undergraduate computer-science course was forty-eight 512-byte blocks, or twenty-four kilobytes. Almost twenty-two million such accounts could be stored on one of the five-hundred-gigabyte hard disks that the MathLAN server currently uses.

I used this space to store programs as I worked on them, but there was not room for me to keep them afterwards, so I deleted the electronic versions as soon as I had paper printouts of the finished programs and their output. Here is a page from one of those printouts. The dot-matrix printer that this particular computer used produced darker text that is nevertheless no easier to read than the output from the printers ten years earlier, because of the distortion of the lower-case letter-forms and the granular appearance of the letters.

I wrote these programs at a video terminal, using a very primitive editor called XED (not to be confused with any software that currently uses the same name). XED was designed to be used on hard-copy terminals rather than video terminals, so it did not continuously display the contents of the editing buffer, but instead reprinted individual lines on command. Editing using XED was a great improvement over retyping from scratch, but otherwise had little to recommend it.

XED featured a command called BACKUP, which created and saved an extra copy of one's file, making it possible to back out of changes and also to keep variants of a program or a handout without having to retype the whole thing.