Giovanni Bajo wrote: > I believe what he meant was that property change should not affect the state of > anything but the *Python*'s object. Then what counts as part of the Python object? If the object is wrapping a C struct from some library, is it okay for a property to change a member of that struct? -- Greg