Re: [gpsd-dev] [gpsd-users] Errors, questions, and FAQ notes

[Eric S. Raymond]

John W. Nicholson <address@hidden>:
> I have not learn how to use gpsd well, but I would like to know how
> to make sure WAAS is working both in Sirf binary and NMEA inside
> gpsd.

gpsd will flip your BU353 into binary mode, but you don't need to care
as it does all the data extraction from these packets automatically.

> The receiver datasheet states the NMEA 0183 Output messages are GGA,
> GSA, GSV, and RMC; along with the optional VTG and GLL.   How can I
> use these option messages with gpsd?

Don't worry about it.  That information is all carried in binary mode;
gpsd will get it and use it.

> The datasheet also states that 48 channel all-in-view tracking, yet
> only a dozen sat are seen with gpsmod. I am not sure if there is a
> need for more fixes in that I am not sure that these are all
> used. If they can be used, how to allow them to be used?

48-channel is marketing-speak.  It means the chip can handle all 48
theoretical GPS channel slots, but consumer-grade chips normally only
handle 12 *simultaneous* channels. This is more than sufficient as you
only need 4 for a 3D fix.

> Using gpsmon with Sirf binary under "Measured Tracker" are the
>  items "Stat  C/N ? SF" what are these four items?

State is some status bits from the Measured Tracker Data packet;
find a SiRF protocol manual to learn how to read those. C/N is
a signal-strength indicator. SF indicates whether the sat is in
use for the current fix.

> Why do I get a "Read error from device" after some time with Sirf
> binary but not NMEA?

I don't know.  I use a SiRF in binary mode abd don't see this.  You may
have some local configuration problem or an issue with your kernel serial
driver - GPSD reveals these with annoying frequency.  Upgrading to a 
newer kernel *might* fix it. Or not.

> I do not know how but, it also seems to work in 9600 bps mode. Is
> this a good thing, will thing be damaged?

It's a good thing. It reduces your fix latency slightly.

> I get the following error message while no gps type device is running :
> 
> $ gpsctl /dev/ttyUSB0 -n 
> gpsctl:ERROR: error 'Multiple subscribers, cannot change control bits on 
> /dev/ttyUSB0.'
> gpsctl:ERROR: /dev/ttyUSB0 mode change to NMEA failed

That invocation is meaningless if there is no device connected, but shouldn't
give that error message - which normally indicates there are multiple (more
than one) client sesions open to gpsd and watching the same device.
 
> What do I need to do to get the GPS in NMEA mode and using WAAS? 

Changing the mode to NMEA won't do anything useful, gpsd changes a SiRF 
device's mode to binary soon after opening it.

You can't do anything specific to make WAAS work.  Either it will or it won't;
see my previous mail.

> How I understand it, the information from 4 gps are merged into one
> data point similar to gpsprof (however as a running total average),
> but broadcast out like single NMEA gpsd stream. I expect them to use
> /dev/ttyUSB0 to ttyUSB4 for devices, but I don't see how they are
> all read/write with only one port. If I use four USB receivers, and
> a hub, how can I do this? Even better what software can I use to do
> this?

When you plug multiple USB GPSes into a machine that has the GPSD 
package installed, it will recognize all of them and deliver JSON
reports from all of them over the sockets that your client applications
open up.

GPSD only manages the devices for you; it's up to your application to
perform merging and data reduction.  I don't know of any packaged
software that does this; you'll almost certainly have to write your own.
-- 
                <a href="http://www.catb.org/~esr/";>Eric S. Raymond</a>