Re: Y2K Fears to Stop Cityrail Trains

[Eddie Oliver <eoliver@efs.mq.edu.au>]

The thing that most amuses me in all this is the assumption that every
computer's clocks are perfectly adjusted to real time. Especially since
it is alleged that some of the biggest problems are with subtle hardware
clocks rather than obvious software-controllable ones, what fantastic
burst of faith is it to assume that every hardware clock is accurate?

Statistically one might assume that there is a fairly wide distribution
of times in the hardware clocks, and therefore there should be a
similarly wide distribution of times when Y2K incidents occur (if they
occur at all, which still seems to me to be an imaginative assumption
driven by people who have made heaps of money by 
"consulting" on the subject).

If a crucial hardware clock thinks it is 0000 in year 00 only when it is
really 0030 in year 00, and the  trains have started running again at
0015, the problem has scarcely been avoided. Moreover the incident could
just as well happen in the middle of the evening peak several weeks
earlier or later, if my perception of the accuracy of some hardware
clocks is accurate. Indeed I doubt that some hardware clocks even know
what year it is.

Eddie Oliver