On Thu, 30 Apr 2015 08:52:09 -0500, Tom Marchant wrote:
>On Thu, 30 Apr 2015 06:31:09 +0000, Vernooij, CP wrote:
>
>>They can claim anything, who is gonna check this and where
>>can I complain after 15b years if my clock appears to be not
>>that accurate then?
>
>Why shouldn't it be that accurate? After all, a second is currently
>defined in terms of atomic clocks.
>
In fact, an average of several, geographically separated for fault
tolerance.
There's a fine metaphysical question here. The meaning of any
physical quantity depends on specifying a process for measuring
it. If you define time as that which is measured by a sundial, the
atomic clock is inferior; at best it measures something else.
But computing an average implies that one can compute a variance
and conclude that pendulum clocks agree with other better than
sundials, so we changed our notion of time from the sundial
convention to pendulums, and subsequently to atomic clocks,
accepting the nuisance of leap seconds.
On Thu, 30 Apr 2015 09:32:05 -0400, Schmeelk, Gregory P. wrote:
>Um, have the taken into account the Einsteinium time dilation that will occur
>as the Andromeda galaxy merges with ours?
>
>I'm just saying :-)
>
But it matters, and it's old stuff:
http://en.wikipedia.org/wiki/Hafele%E2%80%93Keating_experiment
GPS must correct for it.
-- gil
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN
