[LEAPSECS] Coming of age in the solar system

Steve Allen sla at ucolick.org
Sun Sep 5 18:39:43 EDT 2010

On Sun 2010-09-05T12:07:16 -0600, M. Warner Losh hath writ:

> It is also my belief that leap seconds will continue to not exist. As

> Joe Gwinn pointed out, there's two properties of time_t that are

> stronger than UTC or the underlying time scale: it is assumed to be

> monotonically increasing and a number of seconds since the epoch

> formula requires start of day % 86400 == 0.

Leap seconds will not exist in time_t, ever, that's pretty clear,

If the broadcast time scale becomes uniform, then it will provide
"atomic days" of 86400 TAI seconds. But whether that happens under
the name UTC or not, there will be a day every year when a cron job in
the US scheduled at 01:30 happens twice, and a day when a cron job
scheduled at 02:30 never happens.

It seems to me that the notion of "atomic days" in time_t which differ
from civil time is not any more cumbersome than the problems
encountered by programs which don't take care to note the 23 hour
civil day and the 25 hour civil day. And of course those strange days
differ from one country to another, so the full-hour effect on rail
and flight schedules still has to be handled without triggering crashes.

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

More information about the LEAPSECS mailing list