The STMD algorithm now supports parallel reading of all mesh types. Updated the algorithm to remove unused points in each domain so that we reduce memory overhead on unstructured and polydata. Changed the Mutliblock layout so that each node is a multipiece dataset.