Re: Seeking clarifaction of tai64nlocal and leap seconds

[email protected] (Paul Jarc) Fri, 01 Apr 2011 15:10:26 -0400
Newsgroups gmane.comp.djb.syslog,gmane.comp.lib.libtai
Organization What did you have in mind? A short, blunt, human pyramid?
Message-ID <[email protected]>
[Moving to the libtai list, where this is more on topic.]

Brian Reichert <[email protected]> wrote:
> This sentence to me implies:
>
> If you have both right/US/Pacific and US/Pacific timezones, then 
>
> - right/US/Pacific has the leap seconds included
>
> - and hence should be twentyish seconds 'ahead' of US/Pacific.

It's actually that right/US/Pacific assumes that *values returned by
time()* include leap seconds, and US/Pacific assumes that those values
don't include leap seconds.  So if you interpret the same time() value
using each of those timezones, the right/ one will report an *earlier*
time.

Conversely, if you have two machines that both report the same
human-readable time, but one uses a right/ zone and the other doesn't,
then the one with the right/ zone is reporting *greater* raw time()
values than those on the other machine, since the right/ machine's
time() values include leap seconds.


paul