To my knowledge, this closure syntax was a direct Ruby influence mainly because this syntax was also supposed to work like Ruby blocks: `array.each |e| { ... }`

To my knowledge, this closure syntax was a direct Ruby influence mainly because this syntax was also supposed to work like Ruby blocks: `array.each |e| { ... }`. Originally owned and shared closures had a different syntax (`fn~() { ... }` or `fn@() { ... }`)---this was back when Rust had a built-in shared reference, roughly equivalent to `std::sync::Arc` in the modern Rust---and it was pointed out that they should be harmonized at some point [1]. It was that time when people realized that then-loop-only syntax can be generalized into any closures, and this semi-decision got stuck even after the "internal" iterator that accepts a closure has long gone. [1] https://github.com/rust-lang/rust/issues/1864 https://github.com/rust-lang/rust/issues/1864