Some conversions make sense to lose some precision given they go from Float to Int (as in conversions from HSL to RGB). But conversions between two formats that store their components in Float shouldn't have any precision loss. Review all tests to check for those and fix the ones that are assuming loss at the moment.
Some conversions make sense to lose some precision given they go from
FloattoInt(as in conversions from HSL to RGB). But conversions between two formats that store their components inFloatshouldn't have any precision loss. Review all tests to check for those and fix the ones that are assuming loss at the moment.