[SGVLUG] CD permanency

[Terry Hancock]

On Thu, 17 Nov 2005 11:36:22 -0800
Michael Proctor-Smith <mproctor13 at gmail.com> wrote:
> On 11/17/05, Marsden MacRae <mmacrae at earthlink.net> wrote:
> > I can't find (digging through my trash, er "archived LUG
> > threads") any further discussion on this.
> >
> > Can anybody expand on this? We're about to buy more
> > disks, is there some criteria I should use to determine
> > "which disk to buy" for "permanence"? Off list or
> > pointers to informative sites welcomed!
> 
> http://www.mam-a.com/products/gold/archive.html make gold
> substrate CD-R which are supposed to last for a hundreds
> of years (ofcoarse CD-R have not been around for that
> long), So who knows. Gold substrate are supposed to last
> long because silver with corode over time. There are other
> companies that make Gold substrate CD-R.
> 
> Sun light and air can both degrade CD-R so you can look
> into dark nitrogen storage for your archival disks.

The trouble is, CD's haven't been around long enough to
*know* how long they will last (so we can only theorize).

I have read claims that:

   1) CD-Rs are less long-lasting than "pressed" CDs

   2) There are three dye grades of CD-R -- basically
"blue", "silver", and "gold", in order of increasing life
expectancy.  The blue ones are apparently *much* shorter,
while silver and gold aren't all that different.

   3) CD-Rs are very HEAT SENSITIVE, so badly that a CD-R
can become unreadable if stored at high temperatures in as
little as a few months.  So, you want "cool, dry, and dark",
just as with spices.

This and much more detail is available at:
    http://www.cdrfaq.org/faq07.html#S7-5

It should be remembered, however that magnetic media are
also subject to degradation, as are paper and photographic
film.  So, if it really, absolutely, positively has to get
there in a thousand years completely undamaged, then you'd
better look into deep-etching it into metal plates and
sending it to Luna for cold vacuum storage in a lava tube.

Well, okay, that's more for a billion years. ;-)

Note also that if you use a data compression algorithm to
reduce a file by a factor of, say, 5, then store that file 5
times, you can reconstruct the original by a voting
algorithm from the five files, even if they are severely
(but randomly) damaged.  Someday, someone's going to need
to develop that.

Cheers,
Terry

-- 
Terry Hancock (hancock at AnansiSpaceworks.com)
Anansi Spaceworks http://www.AnansiSpaceworks.com