Andreas Rossberg

> In that scenario, a function would have the ability to address all the internal index spaces (tables, globals, functions, etc) in its bytecode. Taken literally, I'm not convinced that's...

There actullay are three interrelated issues. Let's separate them. ### 1. Contref values At first glance, there is no particular problem with handling contref in the embedding API. A contref...

@fgmccabem, that may work for this particular purpose as a (semantically redundant) security measure, with some extra cost and hoops — provided you mean runtime type, since a type index...

@tlively: > It seems that non-castable references should never be able to be passed in from JS to Wasm. Yes, that's what I meant earlier: we couldn't expose these functions...

@tlively: > It's not just the API functions that shouldn't be exposed to JS, though, it's also things as simple as passing and returning contref Yes, absolutely. And that's consistent...

FWIW, here is the complete version of what I suggested at the time: > Here is my early morning take on a very minimal possible design. Let’s say we merely...

The trick here is that we use a tag to enforce the necessary type consistency across places in the code.

If I understand correctly, then I suspect that the "continuation-switched" design won't be type-sound. Note how the typing rules I sketched use the tag to tie the types from switch/suspend...

Wouldn't that essentially be the last row in your table?

@tlively, I'm still a bit lost. Can you write out the typing rules for the handler side, and the reduction and/or desugaring defining execution?