Just to clarify - is the scope of assemble() to avoid big constructors or is there another usecase for it?
That's definitely one of the reasons. Another is to allow the API to evolve over time without necessarily breaking people's code when a new dependency is added to the scenarios. It separates the framework related initialization from the user related initialization. Kind of orthogonal to that is the super() style of initalization. This is very restricting because you can't do much before super, which often leads down the path of having lots of static helper functions and initialization blocks. Can get messy quickly. Though I have to admit, I'd prefer the method to be called something like specify() instead of assemble(). -j
