I wrote (about CD-R longevity):
> However, it should still be quite
> reasonable to expect a substantial portion of that, so even someone as
> paranoid about it as I am should still be able to expect 10-20 years.
I forgot to mention that since CD-R (used as CD-ROM format for data storage)
has three levels of error correction, and any failure mechanisms are
unlikely to result in an abrupt failure (yesterday the media was fine, and
today it won't read at all), you can also assess the quality of the archive
discs at any time. Of course, normally CD-ROM drives don't tell you about
corrected errors, but the better ones have ways to issue SCSI (or ATAPI)
commands to get this sort of information.
So rather than blindly setting some future date to recopy the discs, you
could check up on them every few years, and watch the correctable error
counts.
The error correction consists of three layers of very robust cross-interleaved
Reed-Solomon codes, so it takes a very large number of closely positioned
errors that would be individually correctable before you get an uncorrectable
error.
Received on Tue Jun 13 2000 - 01:11:11 BST
This archive was generated by hypermail 2.3.0
: Fri Oct 10 2014 - 23:33:01 BST