The ORM Foundation

Ken,

The walkthrough in that blog entry (http://blogs.msdn.com/b/adonet/archive/2010/07/19/absolue-beginners-guide-to-entity-framework.aspx) is a model-first approach. He is actually creating an entity model from which the database is then created. It happens that, in this case, there is little difference between the entity model, and some 'ad-hoc classes", but it is a real entity model.

I would be interested in using NORMA to create a conceptual model, from which the entity model would be generated. I would probably use the NORMA-generated DDL to create the database, since I believe it will do a better and more accurate job of creating indexes and such physical-level artifacts.

John Saunders:

but it is a real entity model.

OK. So exactly what is it that qualifies an "entity model" to be a "real" entity model?

What do you see as the measures of "quality" that can be used to test whether an entity model conforms to being "real" or not? 

Ken

I believe he means it's a model made according to the features of the Microsoft "Entity Framework". This extends traditional ER modeling by allowing compound types, which get absorbed into the table entities where they appear, so you can for example define an Address and use it in multiple places as you can in ORM (but not in traditional ER). However, EF is still attribute-oriented, so doesn't have the most important advantages of fact-orientation.