Something I haven't seen in the various articles/presentations about Kotlin switching to a native WASM back end is: what was found insufficient in the existing LLVM back end? LLVM performs some really incredible optimizations in the assembly it generates, and it gets better every day, so I wonder if the Kotlin team can keep up with all the efforts going into LLVM, and whether it's time well spent...
r
russhwolf
10/22/2021, 6:03 PM
I don't know in detail, but I think a part of it was to be able to integrate with the WASM garbage collector instead of Kotlin/Native supplying its own.