Tuesday, 10 March 2009

Obsolete drives; sideways thinking?

I’ve been trying to write this post for ages, but the draft never seemed right, so this is starting all over again, blank sheet.

Many people have a bunch of stuff on media for which it’s hard to find a working drive, whether disk or tape, or punched card or paper tape, or… A feature of the response to my “12 files for Christmas” post was that those who responded have their interesting stuff locked away on such media. There may be other challenges to reading it, but the first is getting the content off the media.

We tend to go all gloom and doom about this. I’ve got stuff on Iomega Jaz cassettes from an earlier Mac, so without a Jaz drive I can’t read it. Might as well chuck it in a bin? Folks have stuff on early Mac 3.5” drives, or Amiga drives, neither of which can be easily read on current systems. And 5.25” or 8” drives are even scarcer in working condition.

Is the only answer to find a working drive on a working computer of the day, the "technology preservation" approach? I remember an ancient engineer scolding me years ago for failure of the imagination, on the subject of disappearing 7-track magnetic tape drives. “Young man” he gruffed, “if you really care about this stuff, lay the tape on your kitchen table, cover it with paper, scatter iron filings across the top, give it a tap, and read the bits off with a magnifying glass!” He didn’t mention that I had to know the relationship between domains and bits, the parity and other features of character encodings, the block structure of the tapes etc; in those days we knew that stuff!

Now I’ve done the maths, and this is NOT feasible with a 3.5”, 2 MByte floppy disk (the iron filings are too big)! It might just be feasible with some of the 8” floppy formats. Different technological approaches (not iron filings, but some other means of making magnetic domains visible, if such means exist) might be feasible for higher densities. In any such case, you would end up with an image of the bits on the disk, in concentric tracks. From here, you have a computational task, or a series of such tasks: identify the tracks, separate into sectors, decode into bytes or characters, decode into directories and file structures, process into files, and now you have something to operate on! Yes, it’s tough to do all that, but you would be able to combine lots of contributions together to do it.

Now, I’m NOT saying that’s the best approach. I AM saying, 8” drives were advanced technology when introduced more than 20 years ago. The requirements for a production 8” drive included high read/write performance (for the time). The requirements now have changed. Performance isn't the issue; scraping every last reliable bit off that drive is!

Today, more than 20 years later, storage engineers in their clean-room high tech environments can build amazingly high performance production drives with previously inconceivable capacities and speeds. But what could you do today with a Masters-level Electronic Engineering lab, some bright students, a few hundred dollars, these ancient media formats, and a much-reduced performance requirement? I don’t care if it takes 10 minutes to read my disk, as long as I can do it!

Is this important? These disks have been stashed away for 20 years, who cares what’s on them? Well, in many cases no-one does. But just think who was using those early drives, and what for. They certainly include authors, poets, scientists, scholars, politicians, philosophers… and many of those people, if not in the first flush of youth then, are moving towards retirement now. Some of these will be candidates for approaches from libraries and archives interested in their “papers”. Previously this meant boxes of paper, photos, diaries etc. Now it includes old media, dropped in the box years ago. Who knows what treasures they may contain? (See the Digital Lives project for examples.)

So, I think there is or will be an emerging interest in these obsolete media and their contents. And at the same time, I think (hope) it would represent an interesting challenge to set students. Perhaps not quite in the same class as building a car to drive across the country on solar power, or robots to play football, but interesting in its own different way.

One of those combined Computing Science and Electronic Engineering schools would be perfect. Would a prize help? Maybe this could factor in something like the Digital Preservation Awards one year? A new kind of Digital Preservation Challenge?


  1. Same idea, different media: The Library of Congress and Berkeley Labs have been working on a process to digitize 78 rpm recordings from a digital image of the disk:


  2. I was about to post the same thing as Perry, this post reminded me of hearing about that LC laser turntable thing. Although at the moment we still have turntables, some old media can't be played without harming it.

    But googling around for this, apparently (putatively) consumer technology now? It's hard to tell how much to trust these guys, or what the cost is:

    But I had imagined LC working with some fancy cutting edge research device, is this really mass produced consumer hardware?


Please note that this blog has a Creative Commons Attribution licence, and that by posting a comment you agree to your comment being published under this licence. You must be registered to comment, but I'm turning off moderation as an experiment.