FreeBSD handles leapsecond correctly

David Benfell benfell at
Mon Jan 2 22:12:14 PST 2006

On Mon, 02 Jan 2006 20:54:00 +0100, Poul-Henning Kamp wrote:
> If there are no more leap seconds, POSIX doesn't have to specify
> how to deal with them.
Isn't this a rather unrealistic position?

You can quote POSIX til you're blue in the face, but leap seconds are
how UTC deals with the slowing of earth's rotation.  And to complain
about this seems about as sensible as complaining about leap days.

Now, if POSIX can't deal with leap seconds, then it seems like you
have a choice.  You can stick to UTC because that's the standard that
underlies civil time.  Or you can stick with POSIX and your system
clock can progressively gain time.

It strikes me that POSIX purity here insists on a deviance from
reality.  And that doesn't strike me as particularly sane.

David Benfell, LCP
benfell at
Resume available at

More information about the freebsd-questions mailing list