This means that, for example, using an integer, such as 4, instead of a float, such as 4.0, will not generate an error when a function has a float type-hinted parameter. Obviously, a float will generate an error when an int is required. I've created a few methods that enforce a float to actually be a float during math operations (and not an integer), by checking each parameter using is_float() and then passing the "validated" parameter to the real type-hinted float math function. Check it out:
github.com/chrispecoraro/PHPStrict