Post #1308496

@simon_brooke @agowa338 @blinry that’s way more ambitious than the runtime I dreamed of; part of the reason for making it one giant FFI is that, even aside from different runtime in-memory type representations, the different semantics and scoping rules make arbitrary translation extremely challenging. Encapsulation seemed to me more tractable. But methods written to be attached to objects of my dream runtime could be easier to manipulate that way