Jason Orendorff
Jason Orendorff
For `enif_send()`, we can just always require an env. I'm not sure what kind of situation would lead a user to want to send a message without having an env...
@hansihe We always supply an env, so I think we're OK.
You're right. We could fix it using thread-local storage; I don't see any other way.
It seems best to code to the documented API. I'll patch it.
If this is something you'd consider, let me know. I can give it a shot.
I think @avandolder is interested in helping. I can help too. The fact that tc39/ecma262#2045 is stalled is a major obstacle. If there's anything I can do to help there,...
I'm not convinced that fixing regenerator 0.13.next makes this web-compatible, though. Won't there be a significant volume of existing regenerator deployments out there, until they all upgrade? What's the half-life...
> Plus the overall algorithm is complicated enough that it seemed wise to normatively specify the intended semantics for the "optimized" case rather than leaving it to each implementation to...
Some tests to include: `if foo && bar {}` - however this is parsed, a test is wise `if foo && bar || baz {}` - this one is tricky;...
OK, how to model this... I like how this PR models the non-chaining case `if let P = E {}`. It produces `(if_expression condition: (let_condition))` and that is great. Now...