Current CControllerGC is useless / incorrect

[Kenji Hata]

Right now we compute the forces needed to be applied by the CControllerGC, but actually never use them.

In CControllerGc.cpp, everywhere where there is: data_->des_force_gc_ = data_->gc_model_.M_gc_ * tmp3 + data_->gc_model_.force_gc_grav_;

should be something like: data_->io_data_->actuators_.force_gc_commanded_ = data_->gc_model_.M_gc_ * tmp3 + data_->gc_model_.force_gc_grav_;

[Kenji Hata]

When we actually use the forces generated by this controller, the robot behaves in an incorrect manner. I have tested this controller on both tao and Dynamics3d and it gives incorrect results - the same. When I control joint 3 and 5 to 1.57 and -1.57 respectively, both Tao and Dynamics3d incorrectly stop at 1.925 and -0.986.

[Kenji Hata]

Fixed the error. You just need a sign change similar to issue #139.

[Samir Menon]

Please outline the exact solution required wrt. latest master.