Re: Leap seconds

[Chris Kuethe]

On Tue, May 21, 2024 at 4:55 AM Roger Oberholtzer
<Roger.Oberholtzer@ramboll.se> wrote:
> Our GPS supplier made a comment that surprised me: the NMEA times contain the 
> 18 leap seconds. Presumably these times would be used unchanged by 
> gpsd/chrony?

Depending on the receiver, you may be able to configure the it to not
emit solutions claiming to have UTC time until the UTC offset is
known.

Section 20.3.3.5.1.6 of IS-GPS-200 says "The 24 MSBs of words six
through nine plus the eight MSBs of word ten in page 18 of subframe 4
shall contain the parameters related to correlating UTC time with GPS
time." (https://www.gps.gov/technical/icwg/IS-GPS-200N.pdf) Those
parameters are transmitted every 12.5 minutes, so time could appear to
be wrong for quite a while after receiver startup.

u-blox receivers are good at giving you information about knowing the
UTC offset: "The information that allows GNSS times to be converted to
the associated UTC times is only transmitted by the GNSS at relatively
infrequent periods. For example GPS transmits UTC(USNO) information
only once every 12.5 minutes. Therefore, if a Time Pulse is configured
to use a variant of UTC time, after a cold start, substantial delays
before the receiver has sufficient information to start outputing the
Time Pulse can be expected."
(https://www.u-blox.com/docs/UBX-13003221)

I've seen other receivers where time just magically jumps once the UTC
offset is known.