In this web content, there is a discussion about signed integers, optional integers, and the implementation of new integer types. Some users express their opinions on the topic, suggesting that two’s complement signed integers with only the sign bit set should be a trap representation and that optional integers should be the same size as regular integers. The conversation also touches on the validation of untrusted bytes, the need for operator. for fancy integer types, and the usage of UDLs for defining values of these types in code. Additionally, there are debates surrounding floating-point numbers, NaN, and the potential implementation of different numeric types.
https://social.belkadan.com/@jrose/statuses/01HNRNHBQY4E5MC37KG14R50P7