Re: [gpsd-dev] Skytraq binary support for Venus838

[Kai Harrekilde-Petersen]

Hi Gary,

Sorry for the late reply, but I was away on a business trip.

On 12-03-2018 23:21, Gary E. Miller wrote:
> Yo Kai!
>
> On Sun, 11 Mar 2018 14:38:14 +0100
> Kai Harrekilde-Petersen <address@hidden> wrote:
>
>
> > I have been hacking away on Gary's Skytraq (Venus6) driver to get it
> > to support the binary mode as well as baudrate and update rate
> > changes (-s and -c) with some success.
> Patches?
Patches are forthcoming. I have a few general patches (like max 
baudrate) which I think would be best to send separately, and then there 
are quite some changes to the Skytraq driver. Maybe it would be best if 
I send things to you for review before delivering a "real" patch?


> > In NMEA mode, the S1216 works out of the box (up to 50Hz fixes!)
> Wow!

Yeah, it is possibly the fastest commercial GPS you can get your hands 
on. I plan to use it as a laptimer/datalogger on my motorcycle, together 
with logging accelerations and rotational angles using the 9-DOF 
MPU-9250 chip. This is quite different from 'normal' use so I'm not sure 
if I'll use gpsd in the end or not.

> > If I turn on all messages (which is a lot), I get some previously
> > un-handled/unknown sentences ("$GPDTM", "$GPGRS", and "$PSTI,003" ),
> > which I promptly made driver_nmea0183.c ignore.
> Sooner or later, soone will want them.

I agree with your point, but since I'm a GPS programming noob, I don't 
know how I could make GPSd use them effectively. I just wanted to stop 
the NMEA driver from complaining about them.

> > It seems like [xc]gps wants the equivalent of GPGSA/GPGSV in order to
> > be happy, and I haven't found a way of getting this data, outside
> > going back to NMEA mode.
> The trick is in the mask.  Look in gpsd_json.c, in function json_data_report().
> The TPV is output when the REPORT_IS mask bit is set.

Perfect, this was the missing bit. I looked through the proto driver, 
and from the comments there, I decided to set both CLEAR_IS and REPORT_IS.
From there I just had to fix an endianess bug on the time-of-week and 
now it is working pretty smoothly.

One question: you've added the sky_msg_NN messages (DC, DD, DE, DF, E0, 
E2, E3) that I don't find in the documentation I have (Venus6: AN0003 
v1.4.19, Venus8: AN0028 v1.4.42) and I don't see my Venus8 emitting 
these messages. Could you send me any documentation you have on these?

Best regards,

Kai