Re: [gpsd-users] Is gpsd supposed to be filtering NMEA sentences it doesn't understand?

[Ed W]

On 04/02/2015 15:12, address@hidden wrote:
> Hello,
>
> $xxXDR is the swiss army knife of NMEA0183.

Sure - seems that way

However, note that the problem I'm bringing up *seems* to be that the 
sentence has a talker id of $YX, not that the sentence is XDR. I think 
you realise that, just want to make sure we don't miss the point and end 
up scheduling something different for post next release which is far 
more wide-reaching than just addressing the talker-id?


> It does not fit well into the ATT group, as as much as i know, voltage and 
> current information of various sensors send are send as xxXDR sentence.

Hmm, I think we need to be guided by real equipment in the field.  I do 
have limited access to equipment, but for example my Airmar 200WX (I 
believe it's also whitelabelled by Furuno) uses both specific sentences 
and also XDR sentences to describe comprehensive roll/pitch numbers 
(both absolute and rates of change). In fact most of the XDR sentences 
it emits are to do with roll/pitch...

(Note that the 300WX docs I referenced seem to imply that there is some 
mythical 300WX device with rain, voltage, current and internal sensors 
that are shown via an XDR sentence, but the 200WX, ie real devices, 
don't have these, just the attitude XDRs)

If anyone were interested in doing a comprehensive validation on the 
200WX I could turn on ALL sentences and wave the thing around and get a 
big sentence log from it?  I'm away on business next week though, so it 
wouldn't be imminently (sorry)


> I think, this sentences should be transformed into an own JSON group, perhaps 
> "SENSOR", and echoed as $IISDR in case of pseudo nmea.

So using only my 200WX device, I would cluster the sentences it emits into:
- Attitude
- Weather (temp, pressure, wind, dew-point calcs)


The author of the ipad app NMEARemote has just implemented support for 
"generic XDR data".  I didn't ask more, but this implies that there are 
other instruments out there emitting that sentence and that real users 
have asked him to implement support?

The implication was that there is some standardisation in the sensor 
naming...?  This might imply that the various fields you see in the 
200WX docs I previously linked will be used more widely by other devices 
also?

Thanks

Ed W