To compound
the problem, the hardware is dying (literally) and (being
proprietary) can=B4t be rebuilt in any equivalent manner.
As Tony has pointed out, early hardware is usually far easier to=20
repair than the modern variety.
On my bench at the moment is a programmable calculator dating from 1972.
It could almost be called a desktop microcomptuer, the architacture is
much the same (16 bit (bit serial) processor, firmware ROMs, RAM,
peripherals).
One chip in the memory control section had failed. Now, I'll admit I
couldn't quickly find a source of the exact same part, but I bought a
fuctional replacemnt (74LS95 as against 7495) from a shop in London. And
apart from the firmware ROMs there is not one component in that machine
that I don';t have a full data sheet for. In therory any part could be
rebuilt.
OK, an official schemtic was not available, but based on my knowledge of
related machines, it took me a couple of weeks (and not working 'flat
out' on that all the time) to produce one.
I think I'll have more luck keeping that machine going [1] than any of
the machine made today
[1] And this is not an isolated example. We can all list dozens of
machine from the 'clcassic' period where just about all replacement chips
are available, where full schematics and service manuals are available,
and so on.
> In some cases the software is physically disintegrating too since, in
> the case of many 8-bit micros from the 1980=B4s, the storage medium was
> cassette tape; a temperamental mechanism at the time, let alone now.
Ah. but that doesn't mean it's got more temperamental with age.
TO be fair, a lot of cassette loading problems were caused by having to
get the volume level just right. I would think (although I've not tried
this, not having suitable hardware) that a modern computer/soundcard
could extract the information from a marginal home computer cassette tape
rather better than the simple interface hardware in the origianl computer
could.
> It=B4s not that no computer innovation took place
in the 1980=B4s, just
> that none of it will be recorded.=20
This sounds a little like something that struck me in May 1986 when I
bought my P850 minicomputer. I, and a couple of friends, realised that
unless somebody did something then 20 years of computer history were
going to vanish -- at the time very few prople were preserving the
once-commonplace onld minis and micros. So we did something.
Fordunately, other people thought likewise and there now is a great group
of people around the world preserving different aspects of classic
computing in different ways (and I am certainly not trying to judge what
is, and is not, valuable here!). And to be honest, I think it's now
unlikely that _any_ machine will be totally forgotten.
Feh. Most of the software of the 80's was written to floppy, an=20
excellent storage medium. Give me a 20-year old 8" floppy any day to=20
a 2-year old 1.44MB 3.5" diskette. How many 80's (not 70's) systems=20
Agreed!
used cassette tape as primary storage? True, the
DC-xxxx carts are=20
starting to have problems, but in most cases, this was used as=20
backup, not a distribution medium. In many cases, the problem is=20
with the aging rubber in the drive, not the medium itself.
I'd be more worried about the various writeable DVD types lasting 20-
30 years.
The other problem is finding something to read it. The old storage drives
were relatively simple and well-docuemtned. Apart from the heads, just
about any part could be made in a good home workshop. Stopage formats
were often well docuemtned, so if there's some way of reading the
magnetic trasitions on the medium, it will be possuible to recover the
data.
But I wouldn't fancy making a DVD drive from scratch (even given the
laser pickup). Many of them are undocuemtned ('they're not worth
repairing' -- maybe not now, but what about in 20 years time when you
can't buy a replacement). In some cases the storage format is
delibearately undocumetned.
To be honest I think it'll be a lot easier to keep a 1970's machine
running than a 2000's machine.
-tony