Re: ✘gpsd .23.2~rc1

[James Browning]

On Tue, Apr 12, 2022 at 5:58 AM Florian Kiera <florian.kiera@logicway.de> wrote:

Yo Gary!

I tested "www/gpsd-client-example-code.adoc" example client on both named versions (3.21 and 3.23.2~rc-1). I added an output in "line 56":

printf("Satellites used: %d\n", gps_data.satellites_used);

and experienced different outcomes. In 3.21 every time the cycle is completed this output gives a valid value (>0). In 3.23.2~rc-1 only every 3rd message gives a "valid" (>0) value for satellites used. Is this intented?

Just checked 3.22 too, works the same as 3.21.

Regards Florian

Am 11.04.22 um 20:46 schrieb Gary E. Miller:

Yo Ralph!

On Mon, 11 Apr 2022 09:11:44 +0200
Ralph Nemitz <nemitz@logicway.de> wrote:

the data.raw file is attached. After "gpsdecode < data.raw" I saw
values at nSat and uSat which are corresponding to cgps but in struct 
gps_data_t are no satellites_visible and satellites_used.

Looks good to me.  The JSON shows the used /seensats and cgps does too
My guess is you are checking for used and seen on reading a TPV insetad
of reading a SKY.

I can't debug your code w/o seeing your code.

Attached is a sample almost as simple as the lat/lon dumper which instead dumps Skyviews.

Among the things it doe is it actually checks if there is a suggestion of a valid Skyview available.

You MUST check that the values are set and SHOULD check that floats are finite.

I did not check that azimuth, elevation, and signal/noise ratio are finite because I want to see how many of you actually do read.

skydump.c
Description: Text Data