> I'll try another writeup. Does anyone know where the original papers for > MVars might be? In my mind, full/empty variables date back to dataflow machines. But I don't know where I got this idea. Stefan