@rmerriam The IHMC definition looks appealing to me. But is there any reason in general to provide the option for sole vs ankle relative coordinates? If I recall correctly, that was only to handle BDI’s interface to the Atlas robot. And why are position/orientation split out into separate fields?
@airballking We could talk multi-arm manipulation messages on this thread as well. I think we should expand to have humanoid_nav_msgs, humanoid_manipulation_msgs, etc.