@pleb @icedquinn @Moon They can't fix floats, they're literally built into the hardware, and changing type conversions would break existing code without much gain

If you give it wacky inputs like true - true it converts them to ints and does exactly what you ask it, subtracts them. C would actually give you the same result, and other languages would throw an error

It either works or breaks somewhere slightly lower down. Only real argument against it applies to any dynamically typed lang