[LEAPSECS] Coming of age in the solar system
seaman at noao.edu
Sun Sep 5 20:08:42 EDT 2010
Some applications require mean solar time.
Some applications require "continuous" time.
Some much smaller number require both.
UTC conveys solar time - and also happens to convey continuous time.
GPS conveys continuous time - and also happens to convey solar time.
In each case the secondary timescale is (not surprisingly) more difficult to get at and requires some mechanism to retrieve a schedule of leap seconds.
Other than the dripping and lazy distaste of the ITU - distaste for the current definition of UTC, distaste for the very existence of GPS, and even distaste for the continued existence of their signature timescale, TAI - what prevents "rendering unto Caesar" in defining applications?
Why can't the applications that require mean solar time use UTC - and deal with the interruptions of leap seconds as the price?
Why can't the applications that require continuous time use GPS? Voilà! No more leap seconds to worry about!
And why can't the very small number that require both - well - use *both* timescales if their developers and users are too offput to manage the contingent retrieval of the leap second schedule that would be needed to make do with one or the other?
It is simply the fact that there are two kinds of time at work. Wishing won't make it otherwise.
More information about the LEAPSECS