OT: Archiving data (LONG)

From: William Donzelli <aw288_at_osfn.org>
Date: Wed May 31 15:42:43 2000

> On the other hand, what happens when/if civilization falls into another
> dark period (world war, nuclear holocaust, whatever-doomsday-scenario)?

Spread out the data - world wide. I've been in computing centers (well,
one) that was designed for a nuclear holocaust. If a war comes, I
think that at least one mirror out in Itchybutt, North Dakoda or
wherever, will survive. It may be isolated, and may be down for some
time, but it will be there.

Of course, reading old TRS-80 files will be the least of your concerns.
 
> Longevity of the medium becomes an issue if you have a bleak outlook for
> our future. You also have to count on future generations having the same
> devotion that our current little group does with regards to archiving old
> software.

There will be. Once the museum folk start to seriously study something
like old computers, they never let go. Never. Just try to think of
something, anything, that was studied in a historical context once, but
not anymore. And, as stuff gets older, the studies get more serious. The
good thing is that the museum folk _have_ started seriously looking into
old computers. The ball, as they say, is rolling...
 
> I'm an optimistic pessismist (in other words, a realist). I don't think
> we can always count on future generations to carry on the tradition and
> continue moving the archives from one RAID network to the next in
> perpetuity.

See the above paragraph. Museum folk are very conservative, very careful,
and look far into the future. For example, a friend on the Pampanito
(WW2 submarine on display in the Bay Area) is very confident that the
work they are doing on the boat will be appreciated in 1000 years.

> And who knows, maybe one day the electricity will run out.

Electricity is of no importance. The information content is. Electricity
may be "obsolete" some day, if superconductors deliver what they promise,
but historical information never goes obsolete. The next medium does not
matter.

> Hence, the need for a longterm physical solution. One that will require
> no maintenance from people. Something that can sit in a cave undiscovered
> for millenia to be uncovered at some future date and leave the future
> discoverers in awe of our primitive technology.

There is nothing wrong with taking "snapshots" of the archives for long
term storage. None at all. But it does have to be practical, and should
not be viewed as the standard to which all such snapshots are made. In
other words, even the medium the snapshots are taken in should evolve,
and use the latest-greatest thing.

> My prefered course of action is to keep it digital until some long lasting
> (near infinity), incredibly high density (megabytes/gigabytes per square
> inch) magnetic medium can be invented.

That sounds great. I would wager that today's high-rel hard disks are
already in the "long lasting incredibly high density (megabytes/gigabytes
per square inch)" category.

William Donzelli
aw288_at_osfn.org
Received on Wed May 31 2000 - 15:42:43 BST