due to param dependencies

Since assertions lead to crashes, we need better failure handling. In all
the cases in this patch, the assert is not required.

All the ones with the task id should be replaced with the module base
class.

Ah yes, and this reduces flash space, since the ASSERT macro will expand to
a printf that contains the source file name.

reduces roughly 130 bytes. Almost enough to pass CI.

Plus some style fixes

- v2.0 in PX4/Firmware (3ec37aa1): https://github.com/mavlink/c_library_v2/commit/95d4cb3b21130d67b251f323482f1c3fe1ce0c6c
    - v2.0 current upstream: https://github.com/mavlink/c_library_v2/commit/163d4c84b9eb58a02cab94d039f54c303977d991
    - Changes: https://github.com/mavlink/c_library_v2/compare/95d4cb3b21130d67b251f323482f1c3fe1ce0c6c...163d4c84b9eb58a02cab94d039f54c303977d991

    163d4c8 2018-04-08 PX4BuildBot - autogenerated headers for rev https://github.com/mavlink/mavlink/tree/07c4fc5f617383707317128e87ec9a0489ed3634

This change fragments MAVLink packets more efficiently and therefore increases the net throughput. This in turn makes the connection significantly more stable and the Crazyflie experience overall more usable.

Signed-off-by: Roman <bapstroman@gmail.com>

excess

This define results in super unreadable code.

With the new position control architercture manual mode and
potition mode run through the same logic and as a result
it had a throttle thershold for smooth takeoff also in
manual mode. This is fixed by by ignoring the threshold for
any mode which doesn't control the climb rate.

The same paramater was used twice, once for scaling the yaw stick
and once for limitng just after. This was useless.

The FlightTaskManual... tasks already had their internal
setpoint member variables. I switched them to use the
architecture with setpoint member variables as it was
implemented the commit before. They simplify a lot.

I realized that instead of using the setpoint setters inline
in real world tasks everyone started to have its own setpoint
member variable and only call the setter in the end for all the
privatly generate setpoints. This makes the setter useless and
therefore I switched to member setpoints in the architecture.
They bring more felxibility which is obviously needed but also
less structure which is the price to pay.

This commit is only done to enable incremental testing of Flighttask.

instantiate flight tasks and the refactored position controller class
only use the new functionality if a temporary parameter is set for testing

individual commit for the indentation because otherwise
the diff gets unreadable. this indentation is made because
afterwards the entire legacy position control functionality
is in an else case.

and make it use matrix library for flight task compatibility

# Conflicts:
#	src/modules/mc_pos_control/mc_pos_control_main.cpp

# Conflicts:
#	src/modules/mc_pos_control/mc_pos_control_main.cpp

all the ones with mc_pos_control need the library, we should use cmake dependencies!

because upstream there needs to be a
common mavlink command definition first
and then it can be easily reenabled again