dcodeIO

> Can't you run the standard set of optimization passes in C API? One can, but only on entire functions. The new ExpressionRunner, on the other hand, can execute a...

> This sounds almost good to me, hmm. > like an API for copying if we don't already have one Now that you mention it, yeah. Indeed, an API like...

I guess the workflow has been automatically disabled quite a while ago, probably due to upstream changes not resulting in any changes to the build artifacts. Triggered the workflow manually...

Hrm, seems there are errors upstream preventing the build from completing ([see](https://github.com/AssemblyScript/wabt.js/actions/runs/7612853526/job/20731361000#step:8:25)). If I'm not mistaken, there were some Emscripten changes a while ago also involving the removal of `writeAsciiToMemory`,...

Iirc the code to determine the version to build lives around here: https://github.com/AssemblyScript/wabt.js/blob/72796586d0e03118e18f5f85278f4742091a847f/scripts/version.js#L43-L46 I guess we could either exclude the specific version, create a dummy tag in this repo, or...

No upstream release yet, but I fixed it otherwise.

Sure, a "live view" of sorts that survives a buffer becoming detached would be ideal, if that's an option. Regarding an event: `change` is fine as well, of course.

@binji That's right, technically this doesn't affect the `Memory` instance, but the views on its `.buffer`. Examples for such views are `HEAP32` etc. in Emscripten. Without an event (or growable...

Related: https://github.com/WebAssembly/design/issues/1271, which discusses a mechanism that would make handlers unnecessary.

Still hoping that iterators can be implemented without allocating dynamic memory, but haven't found a solution yet. If no such solution exists, well, you know, GC.