I'm working on some prime stuff, and I need a squareRoot(of:) method that doesn't rely on floating points. So, I use Newton's method. It's relatively fast and super simple. In doing so, I noticed that the implementation for core integers and what I might write for UIntXL is similar. UIntXL has an update(_:) method for in-place assignments. If all binary integers had this method, then perhaps I could make squareRoot(of:) and friends generic over all binary integer.
I'm working on some prime stuff, and I need a
squareRoot(of:)method that doesn't rely on floating points. So, I use Newton's method. It's relatively fast and super simple. In doing so, I noticed that the implementation for core integers and what I might write for UIntXL is similar. UIntXL has anupdate(_:)method for in-place assignments. If all binary integers had this method, then perhaps I could makesquareRoot(of:)and friends generic over all binary integer.