|
Bernie -> (7/13/2003 1:19:59 AM)
|
[QUOTE]Originally posted by john g
[B]I read it in a usenet newsgroup (so take it for what it is worth) but cdr's and cdrw are still not accepted as archival media by the US government. Life expentancy is less than 10 years. If you will need it sometime in the future, the best choice is still magnetic tape.
Storage conditions play a big part in this as well, I still have Radio Shack model 1 disks that I can still read 25 years later, so if you take care of your disks, you might get better life out of them.
thanks, John. [/B][/QUOTE]
Magnetic media is still not a great archival medium. For one thing, it's self-erasing. How fast this happens depends on how it's stored, but it does happen.
The magazine I used to work for did an experiment one time. We needed to clean out our storage cabinets of software that had been sent to us, and was now gathering dust after having been reviewed. I took several disk cases that were designed to hold 50 5.25 floppies and did the following:
[list]
One case was stuffed full with 100 disks, as tight as I could pack them in there and not bend them
One case had 50 floppies put in it, snug but still with space between them
The last case had only 25 floppies put in, very roomy
50 titles were left in their original store display boxes
[/list]
All of this was put into a wooden cabinet in my basement, with nothing else in it, and the closet was sealed with a padlock. (Each disk case had it's own shelf, and the 50 boxes were simply stacked on the bottom of the cabinet.)
18 months later I opened the cabinet and started checking disks. Of the 100 crammed into one storage case, 87 would not read (87% failure). Of the 50 in the second case, 22 would not read (44% failure). Of the third case of 25, 2 would not read (8% failure). Of the 50 still in thier original boxes, all of them were readable (0% failure). The conclusion we came to was that magnetic media, when tightly packed together, has a tendancy to seek a balanced magnetic field, and will wipe out data doing so.
(Note, no attempt was made to verify the accuracy of the data on the disks that could be read. It was simply a matter of putting in the disk and calling up the directory. If the directory came up, it passed, if not, it failed, so the actual failure rate, in terms of damaged data, is probably higher than stated.)
|
|
|
|
