Update from testing a fix provided by Tom Russo.

The code change for Xastir provided in Tom's "feature-supportglonass"
branch appears to fix the problem of DRAWS gps not working in Xastir. I
am continuing to run this version but everything looks OK.

A statement by Tom explaining the problem follows:

> I think I see exactly what the issue is here.  I am now reading the
> S1216F8-GL documentation at www.skytraq.com.tw/datasheet/S1216V8_v0.9.pdf
> more clearly, and realize that the table of NMEA sentences it supports is
> very different than the other devices documented in the same PDF.  That is, it
> does NOT produce GPGGA and GPRMC sentences, it instead outputs GNGGA and GNRMC
> sentences.
>
> And Xastir doesn't even look at those.

A big thank you to Tom for sleuthing this problem.

/Basil n7nix

Tom Russo <russo@...> writes:

> From the sound of it, since Xastir won't accept the data from DRAWS even
> when reading directly from the serial port, there is something about the NMEA
> sentences that the DRAWS hat is sending that Xastir is not understanding, with
> or without GPSD.  So the very first thing to look at is what those sentences
> are and why they're not being recognized.  Hopefully, capturing the raw serial
> stream for a little while will shed some light on that.
>
> Xastir ignores ALL NMEA strings except GPRMC and GPGGA --- those two contain
> all the information Xastir ever needs.  If DRAWS is sending those and Xastir
> isn't recognizing them, it should be straightforward to figure out why once
> we see what they look like. 
>
> I'm looking at the spec sheet for the s1216f8-gl now, and from what I see
> the GGA sentences and RMC sentences the docs say it sends should be
> completely recognizable by the code in gps.c that is supposed to decode them.
> So this is certainly a puzzle I'm keen to understand.