Building from source without a script

Building from sources on other platforms is not well documented at the moment. If you wish to embark on this adventure, the following packages are required to build mc_rtc:

• CMake >= 3.1
• Boost >= 1.49
• Eigen >= 3.2
• tinyxml2
• GEOS (With C++ bindings)
• LTDL (Not required for Windows users)
• nanomsg
• yaml-cpp
• spdlog >= 1.6.1
• hpp-spline
• SpaceVecAlg
• RBDyn
• eigen-qld
• eigen-quadprog
• eigen-lssol (Optional)
• sch-core
• Tasks
• mc_rtc_data
• state-observation

mc_rtc also has a ROS plugin that enables automated robot’s status publication as ROS topics and facilitate the integration with ROS tools (e.g. RViZ), to build this you will need:

• mc_rtc_msgs

If you wish to get Python bindings you will also need the following:

• Cython >= 0.2
• python-pip
• python-numpy
• python-pytest
• python-coverage
• Eigen3ToPython
• sch-core-python

Additional Python libraries are required to run mc_rtc tools:

• python-git (pip name: GitPython)
• python-pyqt5

The following packages are not required but bring additionnal features:

• ROS
• mc_rtc_ros

If roscpp is available during build then tf2_ros and sensor_msgs are also required. This will add some integration between the ROS framework and mc_rtc (e.g. robot’s state publication) and allow controllers to use ROS functionnalities (provides a ros::NodeHandle instance that can be used anywhere in mc_rtc).

Getting started

Once mc_rtc has been installed, you can jump to the next section.