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

[nemo_outis]

John Hasler <jhasler at newsguy.com> wrote in
news:878vjgz3qq.fsf at thumper.dhh.gt.org: 

> I wrote:
>> [UTC] localizes TAI to this particular planet.
> 
> unruh writes:
>> That makes it sound as if TAI is some God given time.
> 
> There being no god, nothing is god-given.
> 
>> It is not. It was developed and defined in the context of
>> precisely this planet.
> 
> It is based on universal physical constants rather then on
> the motion of a particular object.

Only since 1967 has the second no longer been defined in terms 
of the the motion of the earth.  But it has NOT left that 
legacy behind!

The second his currently defined as: the duration of 
9,192,631,770 periods of the radiation corresponding to the 
transition between the two hyperfine levels of the ground 
state of the caesium 133 atom.

Curious number, that 9,192,631,770 :-)

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.

In other words, while the transition time of the cesium atom 
hyperfine states may be a universal constant, the other 
essential part of the definition of a second, the 
9,192,631,770 periods, most definitely is NOT!  It is 
intimately tied to the earth's rotation - a very, very LOCAL 
matter.

In short, the second is NOT defined exclusively in terms of 
universal physical constants - a big chunk of its definition 
is based on a number chosen to mimic an aspect of a very 
local, very non-universal phenomenon, the earth's rotation.

>> If there is a timekeeping species somewhere else in the
>> universe, I promise you they will NOT use TAI.
> 
> They probably will use something similarly derived from
> physical constants.

And perhaps some arbitrary number which reflects some local 
phenomenon in their neighbourhood?

Regards,