[LEAPSECS] 'right' zone in tzdata on Linux

[Steve Allen]

On Mon 2015-06-22T14:03:03 -0700, Henry Hallam hath writ:
> would return the TAI.  However, it appears to be out by one minute:
>
>   $ TZ='right/UTC' date; date -u
>   Mon Jun 22 21:00:29 UTC 2015
>   Mon Jun 22 21:00:54 UTC 2015

As of today the value of time_t in a system clock intended to use the
"right" zones should have a value 25 seconds larger than the value of
time_t in a system clock intended to use the POSIX zones, just as .
Warner Losch has also explained.

This is because the "right" clock is presumed to have actually
counted every second that was broadcast in radio signals, whereas
the POSIX clock is presumed not to have counted the leap seconds.
That's the way in which "right" is right, it is the only scheme
that has counted every second, some rubber, some SI, since
1970-01-01, but "right" is not a standard in any other sense.

So when the "right" code translates the time_t to a string
it subtracts off those 25 seconds before applying the POSIX
rule for translating time_t to calendar date/time.

> Am I using the 'right' timezone incorrectly, missing something else,
> or is there an error in the TZ database?

Is your system clock set 25 seconds faster than POSIX wants?
That's actually kindof hard, because I know of no standard time
package which will do it.

Pictures of this at
http://www.ucolick.org/~sla/leapsecs/amsci.html

--
Steve Allen                 <sla at ucolick.org>               WGS-84 (GPS)
UCO/Lick Observatory--ISB   Natural Sciences II, Room 165   Lat  +36.99855
1156 High Street            Voice: +1 831 459 3046          Lng -122.06015
Santa Cruz, CA 95064        http://www.ucolick.org/~sla/    Hgt +250 m