Re: Getting the correct UTC time -- leap second issue

From: Steve Allen (sla_at_ucolick.org)
Date: 08/04/04 

Date: 3 Aug 2004 22:10:28 -0700

"Mike Chirico" <mchirico@comcast.net> wrote in message news:<N-idnS3iCqIWApLcRVn-hQ@comcast.com>...
> One has no way of knowing how many leap seconds
> need to be added. Reference the following for a graph:
> http://www.ucolick.org/~sla/leapsecs/dutc.html#atomic.png

> My question is -- does anyone know of any open source packages that will
> give the correct UTC time, both going forward with the best approximation of
> when leap seconds will be added, plus, the exact time for all seconds, going
> back to 1972?

The tables in the referenced web page are the best indication of how large the
difference between atomic time and universal time will grow. The fact that
the tables contain three significantly different projections for the difference
makes it clear that predicting it is effectively impossible. Earth rotation
variations are effectively stochastic over short intervals, and not well
constrained by any current model over long intervals.

Civil time is based on UTC which is a form of mean solar time.
Mean solar time is what society has always used, and it has been
customary throughout history to reset clocks to agree with the sun.

Atomic time, which creates clocks that effectively never need to be
reset, is a new invention.
The two kinds of time are incommensurate, but for most applications
the difference is irrelevant.

It may be that civil time ceases to be based on mean solar time. That
social dilemma is the whole point of the referenced web pages. But
that decision is effectively a political one, not a technical one.
In the meanwhile, the setting of almost all clocks remains dependent on
obtaining the time from someone else who knows what time it is.

It is up to the owner of an individual clock to decide what kind of time
is most suitable. POSIX has chosen the traditional answer of mean solar time.
BSD has chosen the new answer of atomic time. The tz timezone library
allows either choice.

What kind of time do you want your system to use?

Relevant Pages

  • Re: Is "malfunctioning" absolute or relative?
    ... these facts are relative to the frame of reference. ... he is experiencing time dilation. ... that he is moving,m by seeing all those clocks coming in and going out ...
    (sci.physics.relativity)
  • Re: Astronomers Vs Harrison (clock maker) ref Longitude problem
    ... Your introduction of a 'frame of reference' for clocks and axial ... the pace of a clock to axial rotation is evident to all. ...
    (uk.sci.astronomy)
  • Re: Conflicting truths in time dilation aka time dilation debunking 101..
    ... > the emperical fact is the clocks did not keep the same rate. ... different reference frames. ... > The universe is playing with you. ...
    (sci.physics)
  • Re: Climate change
    ... What astronomers know is how the axial rotation with a period of 23 ... And the 24 hour day used on our clocks is a timekeeping convenience ... There is no external reference for a uniform 24 hour day, ... given the fact that the period to which mechanical clocks are ...
    (sci.astro.amateur)
  • Re: What is BaT, exactly?
    ... You obviously seem to think that the existence of one such frame of reference - like the GPS - can make "now to be now ... ...or by clocks that fly through differnet amounts of the Earth's magnetic field and experience completely different turbulence, accelerations and temperature and pressure regimes during their trips? ... interpretations based on previous and similar observations and measurements. ...
    (sci.physics.relativity)