Re: Property Overloading RFC

Derick Rethans wrote:

class. The magic methods are responsibly for: - checking if the property actually "exists"

This is something I wouldn't encourage as one of the virtues of fully dynamic languages is the "lack" of checking this. Otherwise what happens if e.g. one inherits such a class to extend it? Does it have to know about the get/set methods and the necessary declaration of attributes? And as far as "magic" fields are concerned: IMHO the currently available mechanism of checking the property name against a (possibly e.g. from a DB dynamically created) array in the object is sufficient. So personally I'd not add a language feature as it adds complexity with little (or arguable) gain. My $.02, - Chris