It is a very simply form of station keeping. If a boat is
loitering on a waypoint and it gets blown off a distance
past the WAYPOINT_RADIUS it will automaticlly drive back to
the waypoint.
Rover now accepts a new message MAV_CMD_NAV_SET_YAW_SPEED
which has an angle in centidegrees and a speed scale and the rover
will drive based on these inputs.
../../ArduCopter/GCS_Mavlink.cpp: In member function 'void Copter::send_extended_status1(mavlink_channel_t)':
../../ArduCopter/GCS_Mavlink.cpp:281:37: error: suggest parentheses around operand of '!' or change '&' to '&&' or '!' to '~' [-Werror=parentheses]
uint32_t sensors_error_flags = !control_sensors_health & control_sensors_enabled & control_sensors_present;
^
compilation terminated due to -Wfatal-errors.
cc1plus: all warnings being treated as errors
Thanks to Ralph Campbell <ralphcampbell1@gmail.com> for the bug report.
Also supports SET_POSITION_TARGET_LOCAL_NED but for both messages only the position fields are consumed. Support for velocity, acceleration and yaw fields are not included.
That way we avoid recompiling source files that don't depend on vehicles.
Change notes:
- UTILITY_SOURCE_EXTS: moved to ap_library
- Task generators indexes: ap_library handles that now and in a better way
- Use of AP_STLIB_FEATURES: it doesn't make sense anymore, since the stlib
taskgen has empty source list
- Flags and defines: passed down to ap_library through AP_LIBRARIES_OBJECTS_KW
- Set use='mavlink' in AP_LIBRARIES_OBJECTS_KW instead of calls to
bld.ap_stlib().
This will enable users of Rover to plug their sonar's in (or any other
device into the 3.3v ADC) and log the data but not use the data in
navigation and obstacle avoidance.
I have disabled steering disarming for skid steering rovers. Its
perfectly reasonable for a skid steering rover to go hard left on the
spot without any throttle and the user wouldn't want the rover to
disarm during this procedure.
If you disarm from the GCS for a skid steering rover we also ensure
the steering channel is disabled.
Commit 3636b53#diff-e22a85a55f71f1b9b2d3f293dea61368 introduce arming for rover. But in skid steering mode it was only applied on throttle and the rover still pivot.
The patch also correct rover behaviour in case of failsafe or loitering.
GCC 6 has a new warning about misleading indentation:
../../APMrover2/system.cpp: In member function ‘void Rover::set_mode(mode)’:
../../APMrover2/system.cpp:272:5: warning: this ‘if’ clause does not guard... [-Wmisleading-indentation]
if (control_mode == AUTO)
^~
../../APMrover2/system.cpp:275:2: note: ...this statement, but the^Bn latter is misleadingly indented as if it is guarded by the ‘if’
control_mode = mode;
^~~~~~~~~~~~
The issue here is that we are mixing tabs and spaces. Remove tabs and re-indent
the code.
The new function can deal with a variable number of function parameters.
Additionally, I renamed the functions to norm(), because this is the
standard name used in several other projects.
Further to refactor of RC_Channel class which included
adding get_xx set_xx methods, change reads and writes to the public members
to calls to get and set functionsss
old public member(int16_t) get function -> int16_t set function (int16_t)
(expression where c is an object of type RC_Channel)
c.radio_in c.get_radio_in() c.set_radio_in(v)
c.control_in c.get_control_in() c.set_control_in(v)
c.servo_out c.get_servo_out() c.set_servo_out(v)
c.pwm_out c.get_pwm_out() // use existing
c.radio_out c.get_radio_out() c.set_radio_out(v)
c.radio_max c.get_radio_max() c.set_radio_max(v)
c.radio_min c.get_radio_min() c.set_radio_min(v)
c.radio_trim c.get_radio_trim() c.set_radio_trim(v);
c.min_max_configured() // return true if min and max are configured
Because data members of RC_Channels are now private and so cannot be written directly
some overloads are provided in the Plane classes to provide the old functionality
new overload Plane::stick_mix_channel(RC_Channel *channel)
which forwards to the previously existing
void stick_mix_channel(RC_Channel *channel, int16_t &servo_out);
new overload Plane::channel_output_mixer(Rc_Channel* , RC_Channel*)const
which forwards to
(uint8_t mixing_type, int16_t & chan1, int16_t & chan2)const;
Rename functions
RC_Channel_aux::set_radio_trim(Aux_servo_function_t function)
to RC_Channel_aux::set_trim_to_radio_in_for(Aux_servo_function_t function)
RC_Channel_aux::set_servo_out(Aux_servo_function_t function, int16_t value)
to RC_Channel_aux::set_servo_out_for(Aux_servo_function_t function, int16_t value)
Rationale:
RC_Channel is a complicated class, which combines
several functionalities dealing with stick inputs
in pwm and logical units, logical and actual actuator
outputs, unit conversion etc, etc
The intent of this PR is to clarify existing use of
the class. At the basic level it should now be possible
to grep all places where private variable is set by
searching for the set_xx function.
(The wider purpose is to provide a more generic and
logically simpler method of output mixing. This is a small step)
This has no side effects, but since all implementations were basically
the same, move the implementation to GCS_Common and the only part that
adjusts the rate based on which which stream to each individual
GCS_MAVLINK implementation.
This header is used by waf to contain the generated version macros,
particularly using the git hash. For waf it's better to be in a separate
header since it then can keep track of changes on it a trigger
recompilation.
For the make build system, a dummy ap_version.h file has been added in
the missing/ folder so both implementations can co-exist.
Having the version macro in the config.h and consequently in the main
vehicle header means that whenever the version changes we need to
compiler the whole vehicle again. This would not be so bad if we weren't
also appending the git hash in the version. In this case, whenever we
commit to the repository we would need to recompile everything.
Move to a separate header that is include only by its users. Then
instead of compiling everything we will compile just a few files.
I noticed the doco on the Rover discourse page for LOG_BITMASK was
incorrect so I have fixed up a number of doco/comments so they
auto-generate the doco correctly for the wiki. There is no actual
code changes in this commit.
Currently we need to build a different binary for each type of copter frame.
This is a preparation for setting all those binaries to a group called
"copter". We create the groups for the rest of the main products for
consistency.
Fix the following build error on clang:
../../APMrover2/test.cpp:164:4: fatal error: variable 'fail_test' is uninitialized when used here [-Wuninitialized]
fail_test++;
^~~~~~~~~
../../APMrover2/test.cpp:139:19: note: initialize the variable 'fail_test' to silence this warning
uint8_t fail_test;
^
included update_trigger function
added camera trigger precise time mark
detect camera feedback pin status
added support for TRIGGER MSG
corrected according to defines.h
The task generator for the mavlink headers is given a name 'mavlink' and is
added as a dependency to all local static libraries, because all of them use
sources that include mavlink headers. It would be nice to have a good way to
detect sources that use mavlink headers or to declare list of libraries that
use the headers, so that the dependency could be added automatically.
Additionally, there's a minor change in syntax for the task generator: using
strings directly instead of lists.