[Top][All Lists]

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

Re: [gpsd-users] Is gpsd supposed to be filtering NMEA sentences it does

From: Eric S. Raymond
Subject: Re: [gpsd-users] Is gpsd supposed to be filtering NMEA sentences it doesn't understand?
Date: Thu, 5 Feb 2015 07:18:46 -0500
User-agent: Mutt/1.5.23 (2014-03-12)

Ed W <address@hidden>:
> I'm not sure where your referenced list comes from? 

The packet.c source code.

>                      The gpsd web page lists
> "$YX" as being the talker id for "transducer"
> Assuming your list is from the source itself, would it be sensible to sync
> the source with the list of known talker ids on the web page?

If it were "sensible" I'd have done it already.  Adding new leader
prefixes to the packet getter increases the complexity of what is
already (rather unavoidably) the worst single hairball in the
codebase.  Thus I don't do it speculatively - only when there is real,
demonstrated demand.

Adding any *single* leader, however, is not difficult.  I've added $YX,
so $YXXDR should pass through and be visible if you have raw mode enabled.

> >2. GPSD's job is not to reveal NMEA quirks, it's to hide them.
> >Those sentences are a case in point; obviously XDR ought to be
> >translated to a JSON ATT report.
> Actually, be careful.  The XDR seems to be a general purpose way to provide
> transducer data and appears to be the direction NMEA are suggesting vendors
> should implement certain sensor readings.  eg I note several older sentences
> deprecated and referred to using XDR messages

Looking into this, I see that XDR is so underspecified that parsing it 
may be a problem.  There is some hope that we might be able to dispatch on
the sensor name field. But this is absolutely not going to happen before

I've saved the two Airmar PDFs.  You can speed up the implementation by 
doing the following:

1. Transcribe relevant portions of the PDFs into a flat-text description
   of what you want parsed.  Use the format of the NMEA page to describe
   sentences and fields.

2. For each sentence, provide example data that I can use as a
   regression test.
                <a href="";>Eric S. Raymond</a>

reply via email to

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