Yeah, those look really useful for gamedev or simulations, where you need the speed and don't particularly care that it's always the same result...

substr_range is awesome. There's so many times where knowing where in the string the substring is from would have been helpful, especially with parsing. For example, you can calculate where in the input a particular token comes from and even select the full original source that matches a particular pattern by parsing, then calculating what was parsed from the cursor's new position (what's the remainder's range?) and the starting position (what was the range before parsing the pattern?).