Moved the static num_dev_done variable from the bulk and sync read done functions to the master data structure.

Solved a bug in the functions for the sync and bulk operations (the number of servos read was not reaset when the operation was done or there was an error.

Solved a bug in the write function when updating the walk parameters (the ram_offset was used).
The walk move and time parameters are initialized the first time the pre_prodcess function is called.

Used short int's to keep track of the ammount of RAM and EEPROM used to be able to use more than 256 bytes.

Updated the current servo angles with actual feedback or the target angle depending on the servo.

Solved a bug in the USART1 inclusion guards.

Moved the darwin math and kinematics modules to the darwin firmware repo.

See merge request !11

Solved some bugs in the EEPROM data structures.

Removed the motion_pages variable from the library.

See merge request !10

Updated the joint motion module (untested).

Solved a bug in the deletion of an operation: The master index was not reset.

See merge request !9

Solved a bug when re-scanning: previously seen devides were not updated.

Moved the scan function to a different timer channel.
Solved a bug in the operations delete function.

Solved a bug in the motion manager: write data is stored in RAM.
Problems with muliple calls to the setup function(): still not solved.

Added two new variables in the dynamixel manager and motion manager modules: a vector with the present devices or servos.
Solved a bug in the default initialization of the motion module.

Added support ot perform a scan on demand.