A focused walkthrough of Go's numeric type conversion rules, covering integer-to-integer conversions (sign extension and zero extension with truncation), floating-point-to-integer truncation (no rounding, always toward zero), and integer/float-to-float conversions (rounding, not truncation). Includes concrete code examples showing potential data loss and surprising rounding behavior. Also notes that conversions involving floating-point values where the result type cannot represent the value are implementation-dependent, potentially allowing panics on certain CPU architectures.
1 Impression