[ntp:questions] Purpose of a leap second file?

unruh unruh at invalid.ca
Sun Mar 4 20:37:43 UTC 2012

On 2012-03-04, John Hasler <jhasler at newsguy.com> wrote:
> nemo_outis writes:
>> It was specifically chosen, not because of its deep significance as a
>> 'universal physical constant', but specifically to make the 'new'
>> second very close to the old 'earth-rotation' second.
> But not exact.  Can't be: the rotation is not stable.  I'm not
> advocating the elimination of leapseconds: just a cessation of the
> practice of treating them as if they were corrections of errors in the
> atomic clocks.  Let the clocks run on TAI (and so record time stamps) and
> then look up and insert leapseconds as needed for display.

Nor is the measurement of the transition of the Caesium atom stable. In
fact the best clocks are no longer those. No clocks are "stable" They
all have variation and inaccuracies in measurement of them. That atomic
clocks now happen to more stable than the earth's rotation is an
accident of history, and in future there may well be other even better
clocks, and the second will get redefined again. 
If you want your clock to run on TAI, let it. But for most of the
population, time is something used to organize your days. And they will
become more important again as energy becomes so expensive that people
will not have lights on at night, except in special cases. To organize
the day, the sun is pretty important. 

But if you want to measure time with TAI, go ahead. Noone is stopping
you. It is just when you claim that your religion is the one and only
true religion and everyone should comply with the church of the TAI that
people get a bit upset. 

More information about the questions mailing list