Jez Ng

Yeah, I was lazy about implementing another AST traversal to extract out the function declarations. We should do that. This also throws when it shouldn't, because we don't hoist var...

@michaelficarra I actually have [another interpreter](https://github.com/int3/closure-interpreter) that implements hoisting (and several other features) correctly. It does use estraverse, but not escope -- I'll definitely have a look at the latter...

I think promises aren't necessary. We can just use the usual callbacks (like we are currently doing for errCont).

I've experienced a similar issue. It seems that ghc-modi doesn't show errors if a file is `check`ed repeatedly unless its `mtime` changes between checks. And it also has a tendency...

If `ghc-modi` is doing caching, then shouldn't it return the _same_ errors for a file until its `mtime` gets updated? I'm saying that it's not returning any errors at all...

Misc notes for future reference: - Graphics is currently outright disabled because `java.awt.headless` is set to `true` in system_properties. - After toggling it to `false`, the next required step is...

I know it's good practice to split out bugs into different issues, but in the interests of not spamming... here's one more suggestion: Could you expose `str/int/emptyAttr`? They seem like...

Suggestion 4: It should be possible to `sink` single CSS styles directly, rather than editing the entire CSS style list at once. P.S. If you think any of these suggestions...

They aren't used for arm64 either. I believe they're only used for 32-bit archs.

Not fully supported; we still have to go through our optimizations to see if they make any implicit assumptions that break with 037.