As I reported some months back, the word from a committee that studied the
matter is that the data on any recordable CD will outlast the technology if
the CD is stored properly (in case on edge at reasonable temperature and not
in bright light). "Outlast the technology" means that the data will still be
good when you no longer have a device that can read it. A gold CD might last
longer if it's stored on your roof exposed to the elements, but so what?
There is no reason to Zip or otherwise encode data for the purpose of
providing a check on whether or not it is still good. There are multiple
levels of checking already in the data CD format. The basic CD format (even
for music CDs) has a lot of redundancy, error-correction, etc., and then
there is a layer of checking on top of that for data CDs. It is so difficult
to create a CD with data errors on them that people who are making CD
readers (hardware or software) spend a lot of $$ to buy special test disks
that have unrecoverable errors in order to test their devices. If your data
have gotten corrupted, the hardware or software will know it.
Of course, if your CD writer is defective, it can produce a CD that it
or other units can't read. That should be checked from time to time. But if
the CD starts out good, it should stay good.
(Data recording densities on all sorts of media--CDs, hard mag disks,
mag tapes, etc.--have increased tremendously partly because of advances in
making tiny bits and partly because of coding and decoding advances that
have made it unnecessary to get all the bits right. Your CD player has a
chip in it that can do error-checking and correcting calculations that would
have challenged a mainframe 30 years ago.)
The main data-format precaution is to use the most popular,
non-proprietary data format that will accommodate the data you're trying to
store and the future uses you may have for it. If you're really trying to
make an archive that will last for decades, count on recopying the data onto
new-technology media in a new format every couple of decades, unless you
want to keep an ancient computer up and running. -- Mixon
-----------------------
You may "Reply" to the address from which this message was sent,
but note the following permanent addresses for long-term use:
Personal: bmi...@alumni.uchicago.edu
AMCS: edi...@amcs-pubs.org, sa...@amcs-pubs.org
---------------------------------------------------------------------
Visit our website: http://texascavers.com
To unsubscribe, e-mail: texascavers-unsubscr...@texascavers.com
For additional commands, e-mail: texascavers-h...@texascavers.com
