Re: How to avoid 1Hz batching of 10Hz GPS fixes?

[James Browning]

On Oct 16, 2024 02:46, Chris Dew <chris.dew@thorcom.co.uk> wrote:
Hi Gary,

Thanks for your reply.  Here's the information you requested:

what version gpsd,
root@C934E761:~# gpsd --version
gpsd: 3.22 (revision 3.22)

Gpsd 3.22 is unsupported by us. Please update to 3.25 or better yet git HEAD.
gpsd binary from where,
root@C934E761:~# dpkg --list | grep gpsd | head -1
ii  gpsd                                                                     3.22-4+devuan2                     arm64        Global Positioning System - daemon

root@C934E761:~# cat /etc/apt/sources.list | head -1
deb http://deb.devuan.org/merged/chimaera main contrib non-free

what hardware are you running on
iMX8M

what command line to start gpsd,
root@C934E761:~# ps -ef | grep gpsd | head -1
gpsd      3332     1  0 08:10 ?        00:00:07 /usr/sbin/gpsd -n -P /run/gpsd.pid /dev/ttymxc1

serial or USB connection,
serial

how are you measuring "arrival"
By eyeball.  Do you know of a time-stamping gpsd client?

Gpspipe with the -u flag perhaps? 
See if "gpspipe -w" or cgps works better.
No, I see the same batching in each of these.  (I have disabled all other gpsd consumers, with no noticeable effect on the batching.)


Could you suggest a way that I could see whether the traffic from /dev/ttymxc1 into gpsd is also batched?  This might be a ublox issue rather than gpsd.

I would modify gpscat to prepend the formatted result of time.time().