Andreas Rossberg
Andreas Rossberg
This is an interesting discussion, and I think I agree with the general thrust. But if I may, I have two nits about the way the issue is phrased, which...
> 1. Object layout > [...] > The type system of the current proposal doesn't seem to be able to support this layout, i.e. a header followed by array data....
@nilslice, yes, the name change happened. Thanks for pointing out the oversights. OTOH, this proposal has already been merged into the main spec, so that this repo is no longer...
I'm fine either way, will happily accept a PR.
These are fair questions, but for now they are putting the cart three kilometers before the horse. There is a lot to be resolved first. This is a much more...
@littledan, we are considering Wasm implementations with no JavaScipt whatsoever. Some people have already started building them. Enabling that was an explicit design goal from the beginning, and the reason...
@littledan, not sure if that's what you're asking, but a Wasm module can always choose to keep its memory private. In that case, JS has no way of accessing it....
@lukewagner, unfortunately, there are several problems with that. For example: * From the Wasm side, you can only model something as a "call" to an abstract operation if it is...
@jfbastien > nobody seems to know how to properly hook up an axiomatic memory model > with an operational semantics > > Do we need to? > If you don't...
This discussion is focused on memory operations, but they are not what actually matters much in a memory model! The crux of a memory model is the specification of *possible...