[time-nuts] June 30 2015 leap second
hmurray at megapathdsl.net
Sun Jan 11 00:59:25 EST 2015
michael.cook at sfr.fr said:
> I haven’t looked at the raw data, but using the windows apps:
> a Trimble Resolution SMT is showing the leap pending a Motorola UT+ is not.
> lady heather is not
Interesting. I don't see any leap-pending from a Z3801A. (or a KS-24361)
There is another interesting worm in this can... Various software in the
chain from Paris to your PC may filter the leap-pending info.
There was at least one bug in ntpd where it told the kernel to leap-tonight
several months early.
The way it's supposed to work is that ntpd tells the kernel leap-tonight, and
the kernel will add/delete a leap second at midnight. That gives the ntp
servers a whole day to propagate the leap-pending info to lower stratum
servers and another day to turn it off by the next midnight.
The T2 string from HP GPSDOs includes a slot which is +, -, or 0 for insert,
delete, or none. You can ask it when the leap is scheduled, but the code
didn't do that. It just assumed the leap was scheduled for the end of the
current month. The fix was to add a filter to only pay attention to that
flag in June and December.
These are my opinions. I hate spam.
More information about the time-nuts