[LEAPSECS] The Economist

[Tom Van Baak]

> You think that they could get it right....

>

> ITU-T TF.460-6 states :

>

> 2.2 A positive leap-second begins at 23h 59m 60s

>

> and not

>

> " causing clocks set to UTC to display 23:59:59 for two seconds instead 

> of one"


I think he was talking about what clocks actually do, rather than
what the existing ITU spec says should happen.

Wait until June 30 and see how real world clocks handle the
leap second. It turns out, for one reason or another, very few
clocks on the planet actually display 23:59:60 during the leap
second.

By far most ignore it. Many GPS units (especially those with
NMEA output) will use the double 59 method. Ticking quartz
analog clocks use the double 59 or double 00 trick. Some
famous analog clocks (e.g. Big Ben) use a rubber second.
Obviously analog dials are not capable of displaying a 61st
second. Most digital clocks, even if technically capable of
displaying a time of 23:59:60, choose not to.

There are many ways to handle leap seconds. I realize that
only one of them is the "true and correct" way, but since that
method is so awkward designers have come up with many
creative and practical alternatives.

This is part of the problem with leap seconds; there are so
many different implementations.

/tvb