From Eugene Doherty
Belfast
I was very surprised to see the letter from Martin Gregorie about his
experiments on the ability of computers to cope with the next millennium
(Letters, 23 March, p 63). He appears to have taken the same tack as Feedback,
and made the same logical error.
Certainly one can change the date easily, as he did, and most modern
computers will cope with the year 2000. But now try the following: set the clock
to 23.59 on 31 December 1999, wait a minute and see what happens.
You will invariably find the date has reset to 1 January 1980. Now you can
see where the problem will lie as computer clocks, all over the world reset on
the stroke of midnight. Of course it should be easy for someone to work up a
patch to reset the clock back for a pause of n seconds and then reset
it forward by the same interval to hop over midnight, but no one has done it
yet, and think of how many platforms it will have to work on.
Advertisement
The other point this raises is why the date resets to 1980, at least on
PCs—I haven’t tried it on other machines. Would any computer historians
among New Scientist’s readership know this?
Anyway have fun trying this, as I’m sure I’ll not be the only one pointing it
out.
