I’ll take a look into the other ones, for more than 20 years of OO-programming there are very few excellent books.

Thanks a lot
-stephan

I did find it a bit too heavy on the structural aspects and light when it comes to behavior (i.e. a wee bit too data-centric to my taste). But it is a good discussion, lots of good ideas.

To the list of classics above, I would add two books that have had a big influence on my way of thinking about OOD :

1) ‘Object-oriented software construction’ (B.Meyer) for its groundwork and focus on specification.

2) ‘Fundamentals of object-oriented design in UML’ (Page-Jones). It has three parts, the first two are the usual discussion of OO and UML and can safely be skipped. However, the third part (chapters 8 through 14) will totally make it worth your while.

Best regards,
Yanic

Their approach is informed by Peter Coad’s teachings for business component design. Streamlined Modeling gives an OO model that has a fairly direct mapping to a relational model as far as the data goes, but of course with considerable behavior added as well. One of the things that I like about Streamlined Modeling is the *almost* total absence of ambiguity, which means that it’s pretty hard to run into analysis paralysis. (In my limited experience, the folks who dislike Streamlined Modeling the most are the ones who believe that the “best choice” should be left to the designer.)

One of the ways that Streamlined Modeling contradicts Evans is that Evans pre-optimizes his models for a particular application. Streamlined Modeling doesn’t prohibit optimization, but first one develops each part of the model in full detail (at least mentally) and then removes only the bits that will add no value. One really big difference from DDD is that Streamlined Modeling relies heavily on bidirectional associations between objects, so that you can just as readily find all of the parts on an invoice or all of the invoices that included a certain part.