>Hi Marc --
>
>For the sake of brevity, I've chopped off the long thread of replies. Anyone can just hit Previous if they want to read it.
>
>It appears that you are actually arguing the same point I was originally with Jim and John until I realized we were talking apples and oranges (Thanks, JVP, for pointing that out to me).
>

Could be, did not go back far enough. But honestly and with all due respect? They both gave you eighter answers straight out of the manual, or they redefined the problem. And even if you _were_ allone in you situation, which I can assure you, you are not, and their methods would not apply to you, then should they not acknowledge that fact and say so ... in a loud and clear voice? So that we could come to a better definition of what they imply and come to a kind of a taxonomy of cases where sets of rules apply and others do not.

But unless I did not listen well, that is not what they do: they want us to think that it is the only right way where I think that it can be demonstrated that there is no theoretical basis for n-tier as defined, and that in a lot of cases it is not the desirable way.

>As to your points, they are well taken. The complexity issues of n-tier design come into play mainly with physical n-tier systems and not VFP, IMHO. This is the point that I was trying to make.

And mine was, and I think it is crucial in this discussion, that by advising to break down architectures along those lines, in certain (conceivably minority) cases, one may not be optimizing the development effort, missing oportunities and allocating to resources to objectives that may not be desirable.

>I don't have that hard of a time separating the tiers. Sometimes, I would agree, n-tier philosophy has to be compromised to create the more efficient VFP application. For example, in a situation where the application will never grow, why introduce another "tier" by using all views?
>

I did not say it is difficult. I'm saying why do it if it is not on the agenda?

>OTOH, if you have developed good parent classes and frameworks over time (JimB, please validate this point I'm about to make) then the client benefits from that foundation and the "complex" application is actually delivered in less billable time and more bullet-proof.
>

FWIW, I will validate this :).

Thanks and kind regards,

Marc