AP_AHRS_NavEKF doesn't use the rangefinder, it just delegates the
calls to the right estimator.
For libraries/AP_AHRS/examples/AHRS_Test/AHRS_Test.cpp the initialization
order has also been modified to match the order on vehicles. It's more
correct since it passes a reference when EKF2/EKF3 are already
constructed, while thos constructors use a pointer. Ideally these
should be moved to an init() method though, or use a get_instance().
We should never include version.h or ap_version.h headers directly
on a header since this will trigger a complete rebuild of the
codebase when we commit to the repository. The ap_version.h header
is auto-generated containing information from the current commit.
If we include it in a header, every other file that ends up including
that header (directly or indirectly) will need to be rebuilt. No
ccache's cache beats having to do nothing when the header is just
not included.
version.h contains information that is kept on a struct inside
each vehicle. Rather than using the macros from each vehicle,
the getter should be preferred, which returns an AP_FWVersion
referente.
this fixes the OVERRIDE_CHAN functionality to work correctly with the
new function based mixing code
It isn't a perfect match for ArduPilot mixing, but it is very close
for the key control surfaces, including mixed surfaces
the high loop rate of quadplanes led to less than 1cm/loop change in
height, which got truncated to zero. Adjusting height at 10Hz fixes
that.
Thanks to Marco for reporting this!
some people connect USB to allow connections from a companion
computer. The arming check is sufficient to prevent unwanted battery
failsafes when bench testing
the nav controller can think we have already reached the loiter target
if we were last in a LOITER when we switch to RTL. In that case it
would switch to QRTL immediately
found by Leonard (thanks!)
Lucas De Marchi
Randy Mackay
Peter Barker
Andrew Tridgell
Michael du Breuil
Eugene Shamaev
Pierre Kancir