>but I am creating an object model (because I don't seperate data from code) and in the end I get both ,

Yes, but IMHO u start from what seems to be more close to a
logical data model. I just re-looked the first 4 diagrams of
your lecture, and they remind me more a ER diagram than a
OOA one. Don't misunderstand me, i think this is fine; maybe
we will never agree on the terminology, that is, you can call
subtyping - inheritance, or say that relationship among
entities are has-a, uses relations (conections), etc.

>I think the move from the waterfall model is because of RAD rather than
>use of OO - u can build a perfectly valid OO model in a waterfall manner

I don't think so. IMHO RAD is just a kind of tools that let
the developer prototype faster and that actually match
very well with OO design, but in no way is the base for it.
You can use an iterative design process for an application
to be coded using C++, or SmalTalk, and this has nothing to
do with RAD. Indeed Booch suggest that this kind of design
is the base of OO design, and i'm pretty sure he didn't have
in mind RAD when we wrote his book :).

>btw, re : rational.com - did u read Ed Yourdon," Object Oriented System Design - an integrated approach" ?

Nope. But i have some notes of a lecture he gave here in 1994.
Anyway i can't think of other person in the world that can be
more biased toward Structured Design in OO field than Mr. Eddi
Yourdon :).

J. Luis
PS. Damn, here we go again! :)