Re: [gpsd-users] Strange 14s offset with GPS 18x LVC

[Miroslav Lichvar]

On Thu, Oct 20, 2011 at 10:16:28AM -0400, Eric S. Raymond wrote:
> Miroslav Lichvar <address@hidden>:
> > On Mon, Oct 17, 2011 at 01:51:00PM -0700, Gary E. Miller wrote:
> > > Your gpsd likely has a bad value for leap seconds.  It happens. It will
> > > eventually get the right value.
> > 
> > Could gpsd cache the good value or do some sanity checking? Is that a
> > firmware problem? A clock that can randomly jump by 14 seconds is not
> > a very good clock :).
> 
> Knowing what a "good" value is is very, very hard. Effectively impossible.  
> See 
> 
> http://esr.ibiblio.org/?p=2869
> http://esr.ibiblio.org/?p=2882

What about sticking to the value which was reported most often in the
last hour or day? Then allow only one second changes and only at the
end of June and December. I'm still not sure if it's a bug in the GPS
firmware or a design flaw in some protocol.

(btw, my messages seem to be delayed, is this list moderated?)

-- 
Miroslav Lichvar