archiving as opposed to backing up

[Roger Merchberger]

Rumor has it that Tom Jennings may have mentioned these words:
>On Wed, 2004-09-22 at 21:56, Teo Zenios wrote:
>
> >  People always talk about Atlantis because its something mythical
> > to search for, if it was documented 100% nobody would care, it would
> > be like
> > reading 10 years on income tax forms.
>
>Wow, so you're arguing that we should lose things so that those that do
>remain will be more interesting?!

Well, yes! Sometimes I think: "If I lost my wife..." life could be more 
interesting! :-O

> > Do you
> > really want to document every single thing we treasure about a
> > specific
> > machine to the Nth degree so somebody even 20 years from now could
> > just look
> > it up as needed,
>
>Well, yes!

Think of it as "parity." We *try* to save everything available, but some 
will invariably get lost. Hopefully, we've saved enough so it's not 
super-difficult to figure out what didn't survive.

(oh, and for only 20 years, I don't see the importance. *200* years, 
however...[1] How many people are happy that the plans for Babbage's 
analytical engine hadn't gone *poof*?)

> >  or do you really wish something important was forgotten
> > from the archive so that somebody decides to find one, put it
> > together, and
> > find out what the hunk of metal really does?
>
>I find this idea really strange. So we should intentionally leave
>puzzles of the present for the future to decode?!

Sure. Put a couple of crossword / cryptogram books & whatnot in a time 
capsule. However, WRT the classic machines, I agree that we should attempt 
to leave as little to chance as possible.

There is some redundancy that can help here, too. If a fair portion of info 
about OS-9/68K is lost, but a large portion of info about OS-9/6809 
survives, "computer archeologists" could use the existing knowledge to 
(possibly) rebuild a large portion of what was lost. Sure, it's not 
perfect, but at least they'd have access to *some* of the basics, instead 
of starting from scratch.

Laterz,
Roger "Merch" Merchberger

[1] And of course, to reach 200 years, we must first pass 20... So I guess 
20 years isn't important historically, unless too much is already lost by 
then. Then we're already screwed for the next 180...

--
Roger "Merch" Merchberger   | "Profile, don't speculate."
sysadmin, Iceberg Computers |     Daniel J. Bernstein
zmerch at 30below.com          |