SEDA is essentially a set of independent "services" that communicate with each other via queues, which could further be abstracted as message passing.
The actor model is a set of independent functions, which communicate with each other through message passing.
Aren't they essentially equivalent? Am I missing some key difference?
From looking at http://www.eecs.harvard.edu/~mdw/proj/seda/ (archived here), they don't seem to be. SEDA could easily be implemented over actor model, but an actor-based application doesn't need to have anything like SEDA's stages.