jakobkummerow

Another thought: a key driver behind the plan to let Wasm have structural typing is to enable multi-module ecosystems, where modules can inter-operate as long as they define structurally compatible...

Speaking from just one engine's perspective: I think it is extremely unlikely that we will ever use any bits in a pointer for any module-specific special-casing. The marking phase of...

The "closed" update above does not appear to make sense. Rebasing artifact or typo?

I guess it's due to the commits pulled in by the merge: the mentioned commit 66d90a7 does say "Fixes #138" in its description, but was probably referring to some other...

> > Checking its bits, and potentially branching (and maybe missing) would be an entirely additional cost int31ref would bring. So no, not quite for free. > You can't deref...

> I am confused by the suggestion to add direct support for finally at this point. As the OP says, the primary motivation is module size. When the J2CL team...

As the assembly dumps show, the number of Wasm instructions is a very weak predictor for the amount of machine code that'll be generated. Performance-wise, I have no concerns about...

Ah right, good point that an engine could fold `neg(abs(x))` just as easily as it could fold `copysign(x, -C)`.

I can confirm that we recently landed patches that abolished both `funcref

> Is that helpful? It's a list of situations where interval arithmetic may be useful. Nobody argued that interval arithmetic is never useful. (I think some of the specific entries...