How to make a digital preservationist cry

Jan 04 2011 Published by under Research Data

Put your thesis on a 5 1/4" floppy disk. Put the floppy in a floppy plastic pocket. Masking-tape the plastic pocket onto the inside of a hanging-file folder (containing the paper copy).

Leave the folder with the floppy pocket with the floppy disk in a file cabinet.

Do all this in 1985. Do not look at the folder again until 2011.

Somebody pass me a tissue. My eyes are watering here.

12 responses so far

  • rknop says: can always scan and OCR the paper copy, yes? 🙂

  • bsci says:

    For the ignorant among us, is the issue simply that a 26-year-old floppy disk has no data or is there some special trauma evoked by sealing it in a plastic slip for those years?

    • Dorothea says:

      Well, that's a complicated question. (The simplest ones always are!)

      The funny thing about digital preservation is that at base, it's physical preservation. There's always some thing that the bits are on. Just as with analog media (want fragile? try a glass negative!), some digital media are more durable than others -- and durability of the medium aside, there's also a question of whether equipment still exists to read the physical medium. (SyQuest, we hardly knew ye!)

      In both physical and digital preservation, also, auditing is a key part of the system. If you don't know something may be falling apart, you can't fix it. Likewise with environmental controls, though I suppose things could be worse than a file cabinet in a heated/cooled lab.

      But leaving a physical thing with digital information on it alone and unaudited for 25 years is pretty much NEVER going to be a good idea. Not for floppies, not for CDs or DVDs, not for flash drives, not for spare hard drives, not for tape, not for ANYTHING.

      In this specific case, the physical circumstances were fairly grim. 5 1/4" floppies are vulnerable to being bent. The plastic folder this one was in did not protect it from bending; drop the hanging-file folder once, and fuhgeddaboudit. Worse, the way it was masking-taped in made it quite difficult to get the floppy out without bending it. I think I managed (had to razor-blade the tape), but I suspect my care was in vain.

  • Arlenna says:

    See, this post would have my dad feeling triumphant for keeping all of his old computers and associated equipment for all these years. He's always telling us the story about the time he found priceless information on a 3.5" Mac floppy by getting the old box Mac fired up. But we don't want to encourage him.

  • Markk says:

    I am still not getting it. So what if it bent? They could bend. They were bent all the time. They still worked. Is the plastic disk creased or cracked or something? Remember a 5 1/4 in floppy was read by a head in contact with the plastic, so as long as it is intact it will be readable.

    The only issue I see is that it is likely easier to scan in the paper than find a floppy to put the disk in. That is a very interesting point to keep in mind. Paper is universally digitizable with no DRM or odd format .... Just the search methods aren't too up to date. But for things you are archiving - you would have another copy to search on, so paper isn't too bad for 50-100 years. Beyond that is gets bad ...

    • Dorothea says:

      They could bend, but they don't necessarily like to! It's a risk.

      The disk is now undergoing exploratory surgery (so to speak; we're not actually cutting it up, just popping it into a machine to see if there's any viable data on it). The last batch of disks (stored similarly) out of this project were hosed, so I'm not really any more hopeful about this one.

      We're scanning the paper. The real question is whether the disk had anything that wasn't in the paper. We may never know...

  • rknop says:

    "Spin the data".

    That's what Bodhan Paczynski of the OGLE project told me several years ago when I asked him what they did for backups. Ten or more years ago, you could *almost* back up your disks regularly to tapes. But, disk drive technology has taken off, to the point that backing up disk to tapes is impractical in many cases now. OGLE had already figured that out. They just back up from hard drives to other hard drives.

    I think the more general version of this is: as long as the amount of storage we have keeps going up, just keep *all* of your old data around on your current "primary storage" device, and as you move to a new one copy it all over there. It won't take up very much space compared to the new data sets you'll be generating. To back it up, either "get somebody else to do it" (subscribe to some service), or just copy it to a second (or third) live hard drive. If it's the disk (or whatever) you're using all the time, you'll know if the system as a whole goes down. If you want to make sure all the files are still there, then, yes, more complicated auditing is needed. And, as with any backups, you should check every so often to make sure that the backup disk is getting all the files you think it's getting.

  • At least it wasn't an 8.5" floppy?