"Tom Gosse" [EMAIL PROTECTED] wrote:
>  ... I had the chance to listen in on a conversation about which
>  medium was the best for archiving records.  A National
>  Archives worker said that their research showed that in order
>  of durability and longevity it was:
>  *    stone - last forever but not practical.

I believe CSIRO (a government scientific institution in Australia) 
has been looking at etching text onto glass tiles as a means of 
long-term storage.

Of course even if you find a durable physical medium, there's still 
the question of encoding. We have documents that we can't read 
because the script used is undecipherable (I think Linear A is still 
partly undecipherable, and there's a constant debate over how Inka 
quipus should be read).

There's some interesting discussion about how we should mark 
repositories of high-level nuclear waste in such a way that future 
generations who may not share a common language with us can recognize 
them and steer clear. See:

   http://www.ocrwm.doe.gov/factsheets/doeymp0115.shtml
   http://dir.salon.com/story/people/feature/2002/05/10/yucca_mountain/

>  Electronic media - standards and software change so fast it
>  can become unusable in less than a decade ... I wonder if the
>  programs we record today will be viewable in ten years
>  from now never mind a hundred.

Famously, NASA has telemetry data from the Apollo and other programs 
and even technical blueprints for the Saturn V that it can no longer 
read.

As a software developer, the lesson I take from this is that there's 
job security in being a data-format conversion expert. And that 
anyone who creates a format should document the hell out of it 
(presumably in plain text, carved on a slab of granite).

For electronic data, I think the only key to survivability is to keep 
moving forward. Every five years, copy your information to a new 
storage medium and convert it to the most current format for that 
type of data. But there's an obvious problem there if you're dealing 
with 'lossy' encoding schemes such as the ones used for compressed 
video. A few iterations, and - like a sixth-generation copy of a 
cassette tape or a photocopy - you're left with more noise than 
signal.

Angus
-- 
WWW: http://www.raingod.com/angus/          Blog: http://www.disoriented.net/

Reply via email to