Adrian Cole

@bvillanueva-mdsol @jcarres-mdsol fyi.. I think this is the only major library left that doesn't process the longer id. I don't have experience in c# code, or would have raised a...

The key thing here is testing what you're asserting (ex that you fixed the bug as before we weren't encoding strings properly). Here's a test that should cover it. ```java...

I'd recommend using the test I wrote as it would fail before this change, particularly as the wrong type was used. merging code with no tests isn't a good plan,...

@raam86 yeah I would consider this backfilling a unit test vs adding an end-to-end test (which this isn't). It is still valuable. If you want an end-to-end test, you could...

`_start` is already called first (at least in the test I was doing as wazero calls it by default, so before you can call any exports). Perhaps it is due...

ok I should rephrase this as non-js. ex wasi and freestanding cc @hunjixin The nature of entry points in wasm are different than go, and the impedance mismatch is a...

One of the more curious causes of support is that when you compile with the `wasi` target and don't have a main function, it defaults to assuming you want to...

currently, for lack of a `standalone` target, people use `wasi` to export functions in WebAssembly. The host uses those functions directly, passing parameters, as opposed to indirectly via a `_start`.

For context, in zig, this is `freestanding` and rust, `cdylib`. The main request is to stop support cases where people have to remember to add `main` and have confusing errors...

I should also mention that even when some export uses uses WASI (ex. indirectly via fmt.Println), I believe it is better to have a default to backfill main vs have...