At the beginning of a statement, I can see why it would be unclear, but in expression positions, would the meaning be ambiguous?
tompalmerSat 12 Jul 2008
Also, I realize that reflective type access requires the full form (as in acme::SomeType.type.fields). I'm speaking only about the special case of type names standing alone. Or is the concern about the special case causing confusion?
brianSat 12 Jul 2008
I think your proposal is that a type identifier be evaluated as type literal. We can't do that because the grammar relies on the fact that type names without an associated dot operator are declarations (local vars, closures, etc). You could make special cases in specific contexts, but I don't want to do that.
tompalmer Sat 12 Jul 2008
Is there a reason why we require this:
rather than this:
At the beginning of a statement, I can see why it would be unclear, but in expression positions, would the meaning be ambiguous?
tompalmer Sat 12 Jul 2008
Also, I realize that reflective type access requires the full form (as in
acme::SomeType.type.fields). I'm speaking only about the special case of type names standing alone. Or is the concern about the special case causing confusion?brian Sat 12 Jul 2008
I think your proposal is that a type identifier be evaluated as type literal. We can't do that because the grammar relies on the fact that type names without an associated dot operator are declarations (local vars, closures, etc). You could make special cases in specific contexts, but I don't want to do that.
tompalmer Sat 12 Jul 2008
Okay. Just wanted to throw it out there.