pipcet

This is a really interesting read, thank you. I may not be particularly qualified to comment on this, but my outsider's perspective is that wasm as it stands today assumes,...

I'm not sure how much of a win it would be for asmjs (all data has to be encoded as javascript and go through the parser), but wasm would definitely...

I think we'd have to use cell offsets into a global data space rather than pointers for this to work. Or somehow remember which cells need to be relocated and...

Well, the good thing is that things blow up right away if the "wrong" definition is chosen by the compiler—`dup` doesn't work (or at least it didn't for me). If...

Apparently it was longer ago than I thought—8086s apparently used the other result for push %sp. I thought some non-Intel CPUs competing with the 80486 got it wrong. My mistake...

That fix makes it obvious that side effects happen before the deref, so that works for me. I still have a very slight preference for `: sp@ SP @ ;`,...

The asm.js target essentially has separate code and data spaces; it gets around the single memory space limitation by compiling to the data space and emitting code directly. The wasm...

I don't know too much about the non-SpiderMonkey engines, but the current code should work under IonMonkey, the Baseline compiler, and the SpiderMonkey JS interpreter, though I say that without...

> Nothing too hard to do, but a bit fiddly. Okay, I've got something working with nodejs at https://github.com/pipcet/lbForth/tree/asynchronous-js. I'll try to turn that into "official" nodejs support next.

Sounds to me like there's a bug when `ia` extends across a stack page boundary, which would mean this is hard to reproduce—for starters, where the stack page boundaries are...