Re: [gpsd-users] epx, epy values never change

[Chris Kuethe]

Please see "How do I report bugs in GPSD?" in the FAQ
http://catb.org/gpsd/faq.html

It mentions using "git bisect" do a binary search to pin down where
the behavior starts and ends. I don't have an example handy, but I'm
sure it's in the list archives somewhere.

On Thu, Oct 20, 2016 at 11:19 AM, Jeff Loughlin <address@hidden> wrote:
> Running gpsd 3.16 on a Raspberry Pi 3 (Raspbian 4.4.23-v7+)
>
> On an older Raspberry Pi I had gpsd version 3.6 installed and working fine.
>
> I'm now setting up a new RPi 3 to replace it.  I installed gpsd version
> 3.16, built from source. Same GPS unit (Adafruit Ultimate GPS module).  My
> application uses libgps, and everything seems to work fine except that the
> values in gps_data->fix.epx and gps_data->fix.epy never change.  I use those
> values to display an accuracy ellipse on a map, which is important for my
> application.
>
> I can reproduce this with cgps.  When I first start gpsd and run cgps, I see
> the Latitude Err and Longitude Err fields populated with what looks like a
> reasonable value after a few seconds, but it never changes after that.  I
> would expect those values to change drastically when I move the GPS away
> from the window, but they do not.  If I stop and restart gpsd, I get new
> values once, but they never change after the initial setting.  I have let it
> run for days without seeing any change.
>
> Since this was working in the old version, I downloaded the source for
> version 3.6 and built it on the new RPi, and I do not have the issue - the
> epx and epy values change frequently, which is what I expect - so something
> must have changed between 3.6 and 3.16 that causes this to break.  All unit
> tests pass with both versions.  I have not tried any other versions between
> 3.6 and 3.16, but I can try to narrow it down to which specific version
> breaks it, if that will help.
>
> For now I'm up and running on version 3.6, but would prefer to be on the
> latest - I'm willing to go back to the broken version for troubleshooting if
> someone can give me suggestions on what to try or what to look at.
>
>



-- 
GDB has a 'break' feature; why doesn't it have 'fix' too?