Slate concurrency mostly involves the unit of the process, which is conceptually a manager for a single Slate interpreter. Concurrency may be performed within the context of a single process, using stack / control-flow interaction, or across processes using messaging.