Re: [gpsd-dev] ✘$GPGST busted time

[Gary E. Miller]

Yo Eric!

On Tue, 29 Mar 2016 08:52:25 -0400
"Eric S. Raymond" <address@hidden> wrote:

> > {"class":"GST","time":"1970-01-03T16:38:33.000Z","rms":1.300,"major":105.100,"minor":9.100,"orient":-40.900,"lat":0.300,"lon":0.300,"alt":0.600}

> Bug fixed.

Sorry, I can't agree.  

<  The JSON dump function for GST was missing the guard that
> should have stopped it from emitting a timestamp field if no valid
> fix had been seen yet.

Still not perfect.  Notice in test/daemon/skytraq-fix.log.chk lines 17-22:

$GPRMC,232713.000,A,4404.1237969,N,12118.8472448,W,000.0,159.2,280316,,,D*71
$GPVTG,159.2,T,,M,000.0,N,000.0,K,D*07
$GPZDA,232713.000,28,03,2016,00,00*5C
$GPGST,232713.000,1.3,105.1,9.1,-40.9,0.3,0.3,0.6*78
{"class":"TPV","mode":3,"time":"2016-03-28T23:27:13.000Z","ept":0.005,"lat":44.06
8729948,"lon":-121.314120747,"alt":1132.560,"epv":36.800,"track":159.2000,"speed"
:0.000,"climb":0.000,"epc":184.00}
{"class":"GST","time":"1970-01-03T16:38:33.000Z","rms":1.300,"major":105.100,"min
or":9.100,"orient":-40.900,"lat":0.300,"lon":0.300,"alt":0.600}


The TPV has the right time, but not the GST.  They should be the same.

Same error 22 seconds later at the end of the file.

> Not as bad as it sounds - you'd normally trip over this case only in
> the first two seconds of operation, before the cycle detector figures
> out the sentence sequence in the NMEA burst.

I agree not as bad as it sounds, but it looks bad in xgps.
It is not just a startup issue.

RGDS
GARY
---------------------------------------------------------------------------
Gary E. Miller Rellim 109 NW Wilmington Ave., Suite E, Bend, OR 97703
        address@hidden  Tel:+1 541 382 8588

pgpykurv0aZcG.pgp
Description: OpenPGP digital signature