Jon Pretty

Having proposed the syntax, and used it for a very short while, it's actually super logical. And it's *more* visually consistent, isn't it?

For me, it's not that it's illogical, I just find that it adds to my mental parsing load considerably. The same elements, reordered - it requires careful reading which is hard to skim at a glance

You say "reordered", I say "in exactly the same order"... ;)

As an unnecessary visual metaphor perhaps, but instead of "def name args" it's "def (name args) OR (args name args)" which is unusual & novel syntax for a boring feature that every other language has boring syntax for.

I think I'm missing something…the location of the def determines where it is defined, and the syntax between the def and the `=` just resembles the callsite. Any other order would involve mental reordering between definition site and callsite, or defining elsewhere, wouldn't it?

To me, anyway, part of what makes method signatures easy to read is the expected order - def then name then types then args then return type. Very easy to scan. The other thing - why spend novelty dollars on this, when boring obvious syntax is already widely known?

Because the boring syntax requires reordering... def foo(a: T)(b: T2)(c: T3): R looks like http://a.foo (b)(c) at the callsite. I don't really understand why you *wouldn't* choose to write things in the same order at the definition site as the callsite...