Date: Wed, 15 Dec 2010 21:48:57 +0200 (EET)
   From: Vitaly Magerya <vmagerya-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>

   On Wed, 15 Dec 2010, Alex Shinn wrote:
   > The intent was that `current-posix-seconds' can be easily
   > implemented in terms of `current-tai-seconds'.

   So, `current-posix-seconds' are seconds since POSIX epoch, not
   POSIX time as you initially phrased it? In this case, I agree.

My understanding of the present proposals is:

(current-posix-second) evaluates to the number of SI seconds that have
elapsed since 1972-01-01T00:00:00Z, plus 6307200, minus the number of
those seconds that were labelled as leap seconds in UTC.

(seconds-since-utc-epoch) evaluates to the number of SI seconds that
have elapsed since 1972-01-01T00:00:00Z.

(current-tai-seconds) is (+ k (seconds-since-utc-epoch)) for some
constant k that I do not know.