Re: [gpsd-dev] Heads up, GPSD release 3.12 is imminent

[Eric S. Raymond]

Hal Murray <address@hidden>:
> > This is the 48-hour warning.  Get any final fixes in now.
> 
> How many people are testing the latest code with live GPS devices?

Me.  I normally test with a Macx-1 and a generic SiRF-III device.
Both are running fine here and I see PPS from the Macx-1.
 
> I have a setup that is broken.  It's 2 devices feeding gpsd and ntpd watching 
> them.  Of course, I just updated ntpd too, and I hadn't updated gpsd for a 
> long time so I don't have any ideas when it broke.
> 
> Symptoms from ntpd's side:
> ...
>  SHM(0)          .SHM.            0 l  50m   64    0    0.000 -599.273   0.000
>  SHM(1)          .SHM.            0 l  49m   64    0    0.000   -0.051   0.000
>  GPSD_JSON(0)    .GPSD.           0 l    -   64    0    0.000    0.000   0.000
>  SHM(2)          .SHM.            0 l  49m   64    0    0.000   38.154   0.000
>  SHM(3)          .SHM.            0 l    -   64    0    0.000    0.000   0.000
>  GPSD_JSON(1)    .GPSD.           0 l  49m   64    0    0.000 -546.854   0.000
> 
> It worked well enough to get a few samples through, then stopped responding.
> 
> gpsmon shows a blank screen, eventually saying:
> Read error from device

That is worrying.  The NMEA2000 guy was reporting someting like that,
but I could never reproduce it.  What device types?

> Time for sleep.  I'll investigate in the morning.

First thing to try is dropping back to 3.11 to see if it reproduces there.
If it does, running a bisection is easy.
-- 
                <a href="http://www.catb.org/~esr/";>Eric S. Raymond</a>