On 2016-04-10, at 21:03, Alan Altmark wrote: > On Friday, 04/08/2016 at 03:53 GMT, Paul Gilmartin wrote: > >> z/OS is in a conspicuous minority by >> running TOD on (biased) TAI. > > Eh? z Systems CPCs, without regard to OS, are supposed to be set to UTC, > not TAI. UTC is needed in order to get the correct local time by the > addition of the tz offset. > No. Not UTC but TAI biased by -10 seconds.
> If you don't use STP, then you won't be UTC and you won't be TAI. You'll > be in a private time warp. > > http://www.vm.ibm.com/devpages/altmarka/vmleap.html. > Which says: ... Because the TOD clock does not pause for leap seconds, the configured number of leap seconds are added to UTC before the TOD clock is set ... And elsewhere that as of July 1, 2015, the TOD clock is 26 seconds ahead of UTC. Since IERS says that as of that date UTC was 36 seconds behind TAI, TOD was (and remains) 10 seconds behind TAI. You're neglecting what you wrote last year. -- gil
