Can this be compared to something like αcτµαlly pδrταblε εxεcµταblε [1], that makes a single executable run on Linux, MacOS, Windows, FreeBSD, OpenBSD, and NetBSD?
While wasmer seems to support more source languages, it requires running the executable under a WASM virtual machine. Are there other important differences?
There's another blog post by Wasmer where they compare with ape/cosmopolitan.
> How is this approach different? ape is aiming to create one standalone binary that works everywhere (same chipset, x86 as lingua franca). On the other hand, Wasmer aims to have one standalone lingua franca (Wasm), that can target multiple OSes (macOS, Linux, Windows), chipsets (x86, ARM, RISC-V) and programming languages (Python, Rust, Go…).
> On one hand, both ape and Wasmer rely on a “universal binary interface” to interact universally with the Operating System: Cosmopolitan (as a universal libc POSIX interface) in the case of ape, and WASI in the case of Wasmer.
> On the other hand, since Wasmer relies on WASI it can also offer a full-sandboxed execution when running the binary (this means that by default no files or sockets could be accessed or created by the binary unless explicitly allowed, for example).
While wasmer seems to support more source languages, it requires running the executable under a WASM virtual machine. Are there other important differences?
[1] https://justine.lol/ape.html