A. sediba is a recently described species of gracile australopithecine from the site of Malapa in South Africa which dates to approximately 2Ma. A recent issue of Science contains several articles which describe in detail the anatomy of several very complete specimens. These fossils are important because they are so complete, and because their discoverers have suggested that the new species may be directly ancestral to our own genus Homo.
The paper by Zipfel et al describes the foot and ankle bones, including an articulated distal tibia, talus and calcaneus preserved encased in matrix in approximately anatomical position. Other foot specimens include a very complete distal tibia as well some metatarsals. The details of the ankle joint are pretty human-like except for a very robust medial malleolus. The real kicker is the calcaneus, which is quite gracile and lacks a weight bearing lateral plantar process (LPP). These features make the calcaneus look rather like a chimpanzee's. Coupled with the robust medial malleolus, this could mean that A. sediba retained a degree of arboreal climbing competence. I feel sure that the functional implications of this morphology will be heartily debated.
What is especially interesting is that A. afarensis resembles modern humans in these details of the calcaneus and distal tibia. A. afarensis (the species Lucy belongs to) is the gracile australopithecine cousin of A. sediba which lived over a million years earlier and is widely considered the direct ancestor of genus Homo. Regardless of whether or not A. sediba is directly ancestral to Homo or not, this mosaic ankle morphology is interesting and suggests some level of independent evolution of modern human-like pedal morphology.