Clean up muscle init

[Samir Menon]

This looks junky and inconsistent with our latest standards...

/** Initializes the actuators. This involves determining whether the muscle
     * matches the robot etc. It also sets up the Jacobians to be computed etc.
     */
    virtual sBool init(const std::string& arg_name,
        const SRobotParsed *arg_robot,
        const sutil::CMappedList<std::string,SRigidBodyDyn> &arg_rbdtree,
        SActuatorSetMuscle *arg_mset,
        CDynamicsBase *arg_dynamics);

[Samir Menon]

Also need to fix the Muscle init() to match the muscle set

/** Initializes the actuator. This involves determining whether the muscle
   * matches the robot etc. It also sets up the Jacobians to be computed etc.
   */
  sBool CActuatorMuscle::init(const std::string& arg_name,
      const SRobotParsed *arg_robot,
      const SActuatorSetMuscleParsed *arg_msys,
      const sutil::CMappedList<std::string,SRigidBodyDyn> &arg_rbdtree,
      CDynamicsBase *arg_dynamics)

[Samir Menon]

Looks like the reference style of the CActuator*Muscle needs some updates to make it re-entrant and to simplify dependencies. As of now the access pattern is a bit ad-hoc...

[Samir Menon]

Fixed a lot of the data access and init stuff as of 078ea609e9ac3491543a465e9415993365ee5351

[Samir Menon]

Added a basic redis muscle control app as of 08aa5250aa68e4af1478187c4b1fc38ab2315044 in muscle_dev.

[Samir Menon]

Adding dyn type init to redis sim c0f62df4fff074ea7554986a35f1938a615bd93b

[Samir Menon]

Localized an issue. IO data structure doesn't init properly when provided muscle actuator sets. The core issue is that the SActuatorSetBase object doesn't have an init method.

[Samir Menon]

Actuator init done as of 63f286ee4dc212722212dd7dadbcf8f99c6aa687

[Samir Menon]

IO ds init properly as of fc87cbc44d2b7a73514252031de803c8acbea54b

[Toki Migimatsu]

Move init code for sim, visualizer, and controller to init function