What 20th century computer do you think will still be usable 200 years from now?

[CraigD]

All this reminiscing in the ”Fossilised computers” thread brings to mind an old, speculative question: A couple of centuries from now, what, if any, 20th century computer do you think will still be usable?

IMHO, 80386 or later boxes are unlikely choices – too many hot chips and fragile moving parts. My pick is something like my old (but still on-hand) TS2068. Supply it’s plug with something close to 5V DC, and it boots to a programming language. It reads and writes data from essentially any analog audio recorder – 1/4" cassette tape was the usual – you’ve got to press Rewind, FF, Play and Record by hand – outputs an RGV video signal (which a clever person with a 19th century oscilloscope should be able to get a picture from), and (though I don’t intend to attempt to prove it) could, I suspect, be taken apart, buried, dug up, cleaned off, put together, and would still work. Nothing but its keys, power jack spring, and the lid on its cartridge slot moves, and nothing inside, even it’s speedy (3.5 MHz 8 bit) Z80A CPU, gets more than barely warm to the touch.

[Tormod]

I think maybe some of the handhelds, like the Palms, might last, if the LCD screens can survive for centuries. All flash memory and built-in ROM, no movable parts, built-in screen etc.

[Boerseun]

I think most modern computers should be pretty much servicable a hundred or so years from now. There's really very little friction in moving parts like modern-day hard drives, provided it doesn't get a shock or a bump, I think the only bits that won't last would be drives that are open to the elements, like DVD-ROMS, CD-ROMS and stiffy drives. The hard drive and motherboards should last quite a long time.

But then again, in the 1930's they found a bacteria capable of metabolising nylon. Nylon have never before in the history of planet Earth existed anywhere until its creation in a lab in the early 20th century. It was such a handy invention, that it featured very soon pretty much everywhere; from industrial strapping material to faux silk stockings. And, of course, the availability of it made it a prime target for the first organism able to eat it. Pretty soon, I guess, the biggest threat to computers will be bugs! Actual, bona fide living bugs, not the kind your neighbour's antisocial teenage son writes on his computer in the basement. What would be the possibility that some weird bacteria might learn to metabolize silicon? There's so much of it around, that the first animal able to do so would find a rich biome all to itself!

[Shyam Bharath]

Hmmmm.... i am not sure whether other components of the computer will exist but the keyboard might exist (as it did from the typewriter stage...)

Even though we have got new methods of input like mouse, touch screen etc... i still find myself tinkering with the keyboard.

Advanced users try to move away from pointing devices to keyboard shortcuts.

[CraigD]

Originally Posted by Boerseun

I think most modern computers should be pretty much servicable a hundred or so years from now. There's really very little friction in moving parts like modern-day hard drives, provided it doesn't get a shock or a bump, I think the only bits that won't last would be drives that are open to the elements, like DVD-ROMS, CD-ROMS and stiffy drives. The hard drive and motherboards should last quite a long time.

Components in computers have gotten more reliable in the past decades - For example, according to this paper, manufacturers' standards for mean time between failures for hard drives in 1993 were between 100,000 and 250,000 hours (about 11 and 29 years), in 2003, the standard was 600,000 hrs. (68 y), and several current manufacturers advertise “high reliability” MTBF of 1,200,000 hrs (137 y). While it looks from this like expecting a hard drive to survive 2 centuries of being power up is a bit much, what I had in mind was a machine that'd been mostly powered down for a century or two successfully powering up. I'm not sure how long a powered-down disk can be expected to retain it's data, but for the sake of this question, let’s assume that a computer “works” if it can simply be powered up, and some sort of useful program run on it by a human being pushing its buttons or other input devices. A blank hard disc would not, in principle, prevent this.

The critical point-of-failure in nearly all desktop, laptop, and rack-form computers since about 1995 is their CPU. Although modern day CPUs have MTBFs around 1,000,000 hr/1 century, they can self-destruct in hours if their cooling system (heatsink/fan, fluid jacket, etc.) fails fully or partially. The best MTBFs of these components I’ve seen are in the 100,000 hr/10 y range. More, most of them have an unsealed moving part – a mechanical fan – much less robust than the moving parts of a modern hard drive.

My guess as to what would happen if you were to seal the laptop I’m now using in a time capsule addressed to the year 2208, then, in that year, remove it, sand and apply the appropriate voltage to its power jack, and press its little power button, is that, if it would do anything at all, it would display the manufacture’s startup logo, a message about being unable to read its disk, then, in an hour or so, expire with a faint puff of blistered chip silicon as its seized CPU fan sat idle. In the unlikely event I continue to use the machine indefinitely in an ordinary (almost always on but mostly sleeping or hibernating) manner, my guess is that, sometime in the next couple of decades, the CPU or some other high-temperature component will die. Well before 2200, I (or my dutiful, technologically-backward successor) no effort short of founding an antique semiconductor factory will be able to replace the failed component.

My question is, would this apply to every box similar to mine, or would a few be OK? And, as Tormod speculated, what about low-power boxes, like Palms and cellphones, where thermal issues are less significant?

In a sense, what I’m asking is if a particular class of computer is like a radioactive element, with the number of working ones following a half life curve (there will effectively always be some Bi in the universe) , or like an age cohort, where the number of survivors falls to zero at a point only a few times the MTBF (there are no 200-year old human beings)?

[PuGZ]

Mean time between failures - Wikipedia, the free encyclopedia :

MTBF is not to be confused with life expectancy. MTBF is an indication of reliability. A device (e.g. hard drive) with a MTBF of 100,000 hours is more reliable than one with a MTBF of 50,000. However this does not mean the 100,000 hours MTBF HD will last twice as long as the 50,000 MTBF HD. How long the HD will last is entirely dependent on its life expectancy. An 100,000 MTBF HD can have a life expectancy of 2 years while a 50,000 MTBF HD can have a life expectancy of 5 years yet the HD that's expected to break down after 2 years is still considered more reliable then the 5 years one. Using the 100,000 MTBF HD as an example and putting MTBF together with life expectancy, it means the HD system should on average fail once every 100,000 hours provided it is replaced every 2 years.

That ought to clear up a few misconceptions stated in the thread so far. That said, I believe that most twentieth century computers will still be functional two hundred years from now -- provided that they are stored in a vacuum. Given regular use, I can't imagine most lasting more than a quarter of that!

[CraigD]

Originally Posted by PuGZ

Mean time between failures - Wikipedia, the free encyclopedia :
…

That ought to clear up a few misconceptions stated in the thread so far.

Thanks’ PuGZ, for the relevant wiki info. Using MTBF of a whole population to get a sense of the life expectancy of a small subpopulation – the present-day computers that aren’t accidentally or intentionally destroyed in the next 2 centuries – is a rough estimation technique which could surely stand some refinement.

That said, I believe that most twentieth century computers will still be functional two hundred years from now -- provided that they are stored in a vacuum. Given regular use, I can't imagine most lasting more than a quarter of that!

I suspect that storing any sort of printed-circuit board in vacuum is a bad idea, as, if no power was being supplied to it, it would eventually become much colder than its designed temperature range, and if power were being supplied, much hotter. I’ve encountered clearly defunct boards showing distinct cracks in their conductors, which I’m pretty sure were due to thermal expansion and contraction, and suspect that other components with microscopic features might have the same vulnerability to low temperature or changes in temperature.

As I recall, really fancy time capsules use room temperature and pressure pure nitrogen to keep their contents warm and non-reacting.

To clarify my original question, however: it assumes that the computer in question is not specially preserved, just falls into normal disuse, and sits in an ordinary location – a closet, warehouse, etc. – until, after a couple of centuries, a curious person finds it, and tries to get it to work.

As some might recognize or have deduced, this isn’t a completely frivolous or technical question. Much of the information we have about past human history comes from data stores that sat unused and forgotten for centuries – paper or vellum books, clay tablets, etc. Some technologists and historians have speculated that our current “computer age” may, in the far future, appear as a sort of “data blackout”, because much of our electronically stored information will prove irretrievable. Although I suspect future historians will use advanced techniques to recover data from failed storage devices, this is an assumption, not a certainty, so knowing the practical lifetimes of everyday info technology devices has relevance to distant future history.

What serious research I’ve read – such as that promoted by the Long Now Foundation - on preserving records over centuries and millennia mostly seem to conclude that physical storage – paper, stone or metal engravings, etc. – are the best solutions.