OT: Archiving data/video/movies/photos/oral history

From: Eric Smith <eric_at_brouhaha.com>
Date: Mon Jun 5 21:15:30 2000

> William Donzelli <aw288_at_osfn.org> wrote:
> I've made my point a few times over now, and this will be my last word on
> the matter - and I don't want this to sound like a flame, but man, you
> have to get a little faith in what mankind can actually accomplish. That
> "world of tomorrow" and "technology marches on" stuff may actually mean
> something...

"Technology marches on" means new products are better than old ones. For
some definition of "better". However, the appropriate definition of
"better" almost *never* includes the product actually lasting longer.
In high tech gear, this is because longevity is simply not an issue. The
manufacturer that sold you a 4x CD-ROM drive a few years ago knew that:

1) you had no way to know how long it would last
2) 4x CD-ROM drives would be obsolete soon and noone would care about them
3) he wanted you to buy his forthcoming 12x CD-ROM drive

To the manufacturer, it's a net *loss* if the drive lasts a long time.

The countering factor is that the manufacture does NOT want the drive to
fail while it's in the warranty period, because that will cost them money,
and they don't want too many to fail within a few years, because their
reputation would be damaged. So their motivation is to build the products
with the least expensive parts that can be expected to last for about
five years.

To the extent that making a product that will be robust for five years,
the manufacturer will use quality materials and parts. But if a gear
that will last five years costs $0.0012, and a gear that will last twenty
years costs $0.0013, guess which one will be used?

You might say that the miniscule difference doesn't justify the cheaper
part, even when they buy millions of them. But consider that not only
are they buying millions of that part, but millions of all the other
parts in the drive as well. If they save a tiny amount of money on each
part in each unit, the savings add up to a non-trivial number.
Received on Mon Jun 05 2000 - 21:15:30 BST