[ Ref. "Archival Media for Digital Data" ]

Kevin McElhone wrote:

> I have my 58-note catalog (organised by title and by composer) stored
> on 8-inch floppy disks for an I.B.M. main-frame computer.  I have been
> unable to read these disks for 15 years, although I have tried to find
> someone with an old computer to convert them for me.  I cannot bring
> myself to destroy them, but who would want the disks which cannot be
> read?  The paper versions I have are suffering from being copied or
> scanned so many times, but they are at least useful.

First off, I'd highly recommend scanning the paper copies into an OCR
before they decay too much.  That'll allow you to "regenerate them"
onto a format for a modern operating system (OS).  At that point, you
can then edit and upgrade the data easily.  But if you still have data
on the disks of which you have no printout, take the next step...

Try to find _anyone_ with an old 8" system that can read them.  If
that's possible (and the disks haven't demagnetized yet) then you may
still be in luck.  I've found several ways to successfully transfer
files from old systems:

1) Least painful:  If you have both flavors of system in the same place
still running:  Load a file transfer program on both systems, and
connect with a cable.  Several options:

a) Laplink (for more modern systems)

b) KERMIT file transfer software.  KERMIT is available for virtually
every OS under the sun, all the way back to CP/M-80 systems.  There
are ways to bootstrap in a KERMIT load, as well, with a stripped down
version.  (I once literally poked in a KERMIT bootstrap with the CP/M
debugger, to pull the rest of the package in.)

By the way, in addition, there are early versions of ZIP (like PKZIP)
that run on many old OS's as well, which will help you preserve file
dates and times and directory structures as well.


2) For running timesharing systems, type the files out one at a time
on a terminal of some kind that has file capture, and later "replay it"
back into a text editor window on the new system as if you're typing it
in for the first time.  I've successfully done this in a number of
ways, including:

a) Any terminal emulator on a laptop with file capture (like
HyperTerminal)

b) An old teletype ASR-33 with paper tape punch and reader (yes,
I still have one, and for a while virtually ANY computer could talk at
110 baud! :-)

c) Tape recording a modem's tones with a cheap cassette recorder, and
replay them back into the modem later.


3) But if the original operating system is not available (which sounds
like the case for you), you must then find something that can at least
mount the media, and read it sector by sector.  I have in the past
successfully pulled *text* data off of media from incompatible OS's
before that way.  There are some data recovery programs out there as
well that understand old OS file formats.

I've even gone as far as to read and store each sector in a "dump"
format, then painfully patch the sectors back together with a text
editor.  (Not recommended though, except in "extreme wish or need"
cases. :-)

So, what is the data storage format?  In other words, what program
wrote the files originally?  Are they simply text files, or was the
data created by an old database program that may have its own internal
format?  (If the latter, which one, and which revision of the software?)


> What does this say about digital music storage in the future?

In the USA, our Library of Congress is fast becoming a "historical
reading device" site.  Microfiche, various old computer systems, early
laser disc players, etc., all have to be maintained unless and until
the archived data can be transferred to another format.  If your data
is in some obscure format, contact them.  They may be able to offer
other suggestions for extraction.

Keith McClary