Deprecating MultiBlock datasets

I am not entirely sure what you mean. Is there a specific aspect of multi-block datasets and parallel-processing that you’re interested in?