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
And Xastir doesn't even look at those.
A big thank you to Tom for sleuthing this problem.
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.