Yeah, from a blank slate that's the rule we ought to have had all along. For source compatibility, I'm wondering how we maintain the existing syntax that works with the semantics the user expects.
@beccadax, what is the status of this effort? Is it far enough along to start a pitch or review?
edit: CC @moiseev @ravikandhadai
I think I can write up a full proposal for a pitch, and dust off the implementation while the pitch thread is running so we can proceed from there
6 Likes
Hi @beccadax This proposal looks great. Just wondering if it was possible to converge on the choice of closures. Are you planning to use a closure-free design?
I am; the current implementation uses a new AST node instead of a closure. The pitch thread I was talking about, with a draft proposal detailing the design, has since been posted: