From: ts <decoux@...>
Date: 2000-11-09T16:15:06+09:00
Subject: [ruby-talk:6193] Re: Time.local bug?

>>>>> "h" == hal9000  <hal9000@hypermetrics.com> writes:

h> Merci beaucoups, M'sieur Decoux. That rhymes, no? :)

 :-)

h> 1. What is your source for this?
h> 2. Are these all the leap seconds so far?

   ftp://maia.usno.navy.mil/ser7/tai-utc.dat  : all leap seconds since 1961 
   ftp://maia.usno.navy.mil/ser7/deltat.preds : predictions

 look also at 

 http://hpiers.obspm.fr/
 http://maia.usno.navy.mil/

h> 3. Is there a reliable way of determining when they will be
h> inserted in the future? They seem to be inserted every 18
h> months in recent years...

 See ftp://hpiers.obspm.fr/iers/bul/bulc/BULLETINC.GUIDE

 You can have a leap second at end of June, December, March or September

 The goal of IERS is to have 'abs(UT1 - UTC) < 0.9s' where
  UT1 = non-uniform time based on the Earth's rotation
  UTC = Coordinated Universal Time (an atomic time)

 If 2 leap seconds are introduced, this mean that the difference between
 UT1 and UTC was greater than 1 second *in 3 months*. Perhaps, rather than
 introduce 2 leap seconds, it's better to change the atomic clock :-)


Guy Decoux