Alon Zakai

Looks like @brson (hi!) created that crate - @brson what are your plans there?

I suppose we could optimize this when `trapsNeverHappens` is set. Then we can assume that doesn't trap, and the write is redundant. (edit: ah, I see that was already mentioned...

Isn't it ``` Type(i32) / / \ \ Global("g1", i32) Global("g2", i32) Literal(1) Literal(2) \ \ / / None ``` ? The union of a global and a literal is...

Ah, there might be bugs in `haveIntersection` and `isSubContents` then. But `intersect()` explicitly says it isn't ready for `Global`s yet, https://github.com/WebAssembly/binaryen/blob/6e7a4a0507986d067550911a3bac59ce76f8453c/src/ir/possible-contents.cpp#L145-L147 edit: I guess it only checks that on `other`,...

Does `wasm-dis` not assume all features are present, so that it can always print a module? If so, isn't it an option to ignore the difference between the from-text and...

Oh, you're right @martianboy ... what confused me is that `wasm-dis` works on a file using features but without a features section. It works because we don't run the validator......

I think it would be better to enable them all. My mental model is that `wasm-dis` can disassemble any input module, no matter the features, so it should "just work"....

@tlively you merged the PR that closed this. Personally I'd prefer the other option, but I don't feel strongly. But as mentioned above I think that PR plus fixing validation...

`wasm-dis` not validating is a bug, I think :smile: I guess we could redefine it as a feature (/workaround), but I suspect it will hurt us eventually when we are...

Thinking about this, is it not safe for strings, at least how we do the lifting? For strings, if you start with imported JS strings, then lift to stringref, then...