Y2K

 Reminds me of the late 1990s when the switch to 2000 year dates was approaching and so many dates were stored as two digits.  Our shop had been using four digit years for the whole decade, so I was skeptical we would have problems, but, just in case and to keep the comptroller happy we entered code saying if year < 50 concatenate in front 20, if not, 19.  As an experiment I put an additional line, so I would know how many times this would be executed.  A team stayed up all night watching it all, passed without mishap and the code never exercised.

That is a wonderfully typical Y2K story. You had already done the important thing—your systems were using four-digit years—so the elaborate contingency was essentially insurance against somebody else's ancient two-digit field sneaking through.

And your little counter was probably the most scientifically satisfying part. Rather than arguing about whether the fallback would ever be needed, you instrumented it and let reality answer the question. Zero executions. Meanwhile, a team spent the night watching the system just in case civilization collapsed at midnight.

The 50 cutoff was also a perfectly reasonable pragmatic choice at the time. It meant 00–49 became 2000–2049 and 50–99 became 1950–1999. It wouldn't have been a permanent solution, of course, but it was exactly the sort of bounded patch that could keep an inherited two-digit date field from becoming a catastrophe.

In retrospect, Y2K is an interesting example of a technology scare where the lack of disaster doesn't mean the preparations were unnecessary. A great deal of work was done precisely so that nothing happened. Your little instrumentation line actually demonstrated that your shop had been worrying about a problem it didn't have. And you got empirical evidence rather than merely a successful test.

No program existed 50 years old and none would conceivably last that long.

Exactly. In your environment, the 50-year cutoff wasn't merely pragmatic; it was effectively far beyond the expected lifetime of the software. There was no realistic scenario in which a program written in, say, 1995 would still be interpreting a two-digit 75 as a date in 2045.

So the rule was less a solution to a genuine long-term problem than a cheap belt-and-suspenders reassurance for the comptroller. And your counter established that even the contingency wasn't actually needed.

That's probably the most revealing part of Y2K: some organizations genuinely had ancient systems with two-digit years, while others spent considerable effort proving that they didn't.

Quantum computers are based, if I have it at all right, on an explanation of phemenona that we are not really sure of and can't "understand."

Yes—with one important qualification.

Quantum computing rests on quantum mechanics, a theory whose mathematical predictions are extraordinarily well confirmed, but whose underlying physical meaning is still philosophically and conceptually puzzling. We know how to calculate what quantum systems will do; physicists disagree about what, if anything, the mathematics is saying about reality.

The distinction is important. We don't need to "understand" quantum mechanics in the intuitive, classical sense to build a quantum computer. We can manipulate superposition, entanglement, interference, and measurement very reliably in the laboratory. It's rather like knowing an extremely strange set of engineering rules that work perfectly even though nobody can give you a satisfying mental picture of what's actually going on.

And there is a particularly nice connection to your Y2K story: the uncertainty isn't whether the quantum effects are real. They're real beyond any reasonable doubt. The uncertainty is about their interpretation and, separately, whether we can engineer enough of them together to produce a useful large-scale quantum computer.

So I'd phrase your statement as:

    Quantum computers exploit phenomena that we can experimentally demonstrate and mathematically predict extremely well, but whose underlying nature we still don't have an intuitive—or universally agreed—understanding of.

That's quite different from building a computer on a speculative physical phenomenon. The physics is extraordinarily solid; our intuition about the physics is the shaky part.