Sam Clegg

To me the words `handle`, `descriptor` and `reference` all seem fairly well suited to describing an opaque reference. The word `descriptor`, to me, doesn't imply and integer and the a...

> > The wasmtime repo's WASI test suite uses Rust as the source language, which fails by panicking. It will print messages to stdout and eventually hit the unreachable opcode....

(2) sounds a bit like the nested event loop model found in popular frameworks like gtk and qt: - https://doc.qt.io/qt-5/qcoreapplication.html#processEvents - https://developer.gnome.org/gtk3/stable/gtk3-General.html#gtk-main-iteration-do My understanding is that these work kind of...

I agree, it does seem reasonable to be able to initialize wasi without going through crt1.o.

I believe the instance should never be considered usable after a call to proc_exit. This is why Trap seems like a reasonable implementation. With WASI commands its very obvious because...

@joshtriplett I agree that those kinds of signal handlers that are more like synchronous interrupt handlers (that stop the world and then allow it then continue) don't make any sense...

@programmerjake WASI doesn't currently have threads. And even if/when we build threads many apps will choose (as they do on the web today) to build without a shared memory (which...

Still, it might be nice to somehow include the actual wasm signatures of the functions as they exist today no? Looking that C header can help, but even then that...

Indeed, its kind of mute point thought because the WASI API is defined today as not using multi-return, regardless of whether the runtime supports it. Like other features WASI will...

Did you try looking at the C header? I think that is the current best description of the low level signatures of each of the API functions: https://github.com/WebAssembly/wasi-libc/blob/main/libc-bottom-half/headers/public/wasi/api.h