gpsd-users
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [gpsd-users] possible week rollover problem


From: Gary E. Miller
Subject: Re: [gpsd-users] possible week rollover problem
Date: Tue, 25 Aug 2015 11:41:43 -0700

Yo Paul!

On Tue, 25 Aug 2015 14:23:49 -0400
Paul Fox <address@hidden> wrote:

>  > > what i've observed:
>  > What we really need is gpsd raw logs, or raw GPS output that we
>  > can look at.
> 
> okay.  i'll figure out how to enbale that, so i can quickly turn it on
> if/when the problem occurs.  can gpsd raw logging can be enabled on
> the fly?  or do i have to change systemd's gpsd conf file to add the
> -D option?

I can't speak to systemd, but you will need to restart gpsd with something
like -D6.

Or you could put the GPS in NMEA mode and just 'cat < /dev/ttyUSB0' to
get the NMEA.

>  > Pretty much all GPS can lose the GPS epoch.  Usually when the
>  > internal GPS battery is dead, but your units sound to new or
>  > that.  But most GPS can report bad time until they have had a good
>  > sky view for 30 minutes.
> 
> in the "weekend trip" scenario i reported, the 2035 date persisted
> over the course of two days of clear sky and good fixes.  i believe
> i was using the S4 device.

Ouch.  At least not a glitch, plenty of time to capture data.

>  > Yup, that would be helpful.  Also, when reporting gpsmon results,
>  > you need to say whether it was in standalone or slave mode.  In
>  > standalone mode gpsmon talks directly to the GPS, in slave mode it
>  > just reports what gpsd is seeing.
> 
> i didn't realize gpsmon could run standalone -- it was definitely
> talking to gpsd in this case.  i'll try connecting directly next time
> it happens.  (the usage message skips that case, btw.)

That is because gpsmon autodetects mode, no need for an option.  If
gpsd is running then gpsmon is a slave.  If gpsd is not running then
gpsmon connects directly.

The man page coovers the topic.

> is the "Time:" reported by gpsmon the gps time?  i'd have assumed so,
> except that it definitely didn't match the results gpsd was giving
> via gpspipe, or to my application.

I'm not that familiar with how the gpsmon SiRF decoder works, but it looks
to me like gpsmon reports the time by sentence type, not one overall time.

So it could be that one sentence type is not being decoded properly.

With the raw data directly from the GPS I can run in gpsfake mode
and try to duplicate what you see.  Check out gpspipe -R mode, super raw
mode.

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

Attachment: pgpCaPiW9KvHH.pgp
Description: OpenPGP digital signature


reply via email to

[Prev in Thread] Current Thread [Next in Thread]