David,

I agree that we disagree.

The original posting highlighted the problem.

The problem is supported by an article in FP Advisor of Feb/98 titled "Dangers of Composite Development".

But what the Hey, eh!

Cheers,
Jim N

>Jim,
>
>I disagree that there's any problem in the VFP object containership hierarchy model. What you are calling a contract is just the defined public interface of an object. For objects to successfully interact they must know each other's interface. And that interface is based on a myriad of design choices.
>
>>Making a "contract between objects" has to be just some fancy terminology/footwork to get around this **problem** of VFP (let me call it the 'object addressability problem'. It is way beyond my limited capabilities, but I have to believe that it *can* be solved elegantly if the right heads are applied to its resolution.
>>
>>And David, I thought that the whole premise of both Windows and Objects *IS* the passing of messages to/fro. I cannot see how this would "break" encapsulation.
>>
>>I can imagine that frameworks will be significantly different once the 'object addressability problem' is resolved. And so will classes generally, as well as their overall usefulness.