On Jan 1, 12:38 pm, Evandro Menezes <[email protected]> wrote:
> On Dec 31 2008, 8:49 pm, [email protected] wrote:
>
>
>
> > So, basically I'm manually adjusting the nmea second reading forward
> > by 1.
>
> > The SHM(0) gps is still fudged to account for normal lag w/ the time
> > given on qnan. Here's my conf with the modified gpsd:
> > #Garmin GPS 18x LVC 0
> > server 127.127.28.0 minpoll 4 noselect  #Local GPS serial
> > fudge 127.127.28.0 time1 -0.420 refid GPSa
>
> How is this different from leaving the source code intact and fudging
> the reference by +0.580?
>
> TIA

Evandro,
because gpsd (and apparently LinuxPPS) have checks that lock it in to
what it thinks is the 0 second boundary. I've tried to fudge it
forward to the correct 0 second, but gpsd gives the error "not in
locking range: 660123." So, it thinks it's 600ms away from the correct
locking area, when it's actually ~0ms away. It's calculated the wrong
second. There was a long thread in the gpsd mailing list where 2
developers were arguing on how to avoid this, and I think they either
took it out completely or made some fix because the code didn't have
the offset anymore.
If you are using gpsd, and you set the clock exactly 1 second ahead
while fudging nmea 1 second ahead with it, gpsd *shouldn't* send the
pulse data because it will claim it's out of locking range.

_______________________________________________
questions mailing list
[email protected]
https://lists.ntp.org/mailman/listinfo/questions

Reply via email to