[LEAPSECS] An example

[Warner Losh]

  On 11/02/2010 14:27, Steve Allen wrote:

> On Tue 2010-11-02T20:22:29 +0000, Poul-Henning Kamp hath writ:

>> It admits no such thing, and for all I know, and despite your

>> complaints about the lack of an "essen" unit, that there is more

>> than one definition of a second in use at present ?

> Section A.4.15

> http://www.opengroup.org/onlinepubs/9699919799/xrat/V4_xbd_chap04.html

>

>      "Note that as a practical consequence of this, the length of a

>      second as measured by some external standard is not specified."

>

> Is this not the POSIX standard?

The same section says there are exactly 86400 seconds in a day.  The 
standard leaves to other standards to define what the elapsed time of a 
second is, and also leaves open the door for errors in time keeping that 
happen in any OS where some seconds are a little shorter and some a 
little longer than others due to temperature changes in resonators, ntp 
steering, etc.   It does not mandate the maximum allowed error in 
timekeeping.  That's very different than saying that seconds are mean 
solar seconds.  It is clear they are SI seconds because time is based on 
UTC (also explicitly stated in section a.4.15).

Warner