I have taken to describing AOMs using the phrase “Code as Data”. Here is my reasoning. In an Adaptive Object Model one uses meta-data to describe objects and, more importantly, the relationships between objects. Once these relationships are captured in meta-data, it then is very easy to think abstractly about the design and to play out “what if” scenarios. Accidental complexity becomes obvious as redundancies in the object model become more apparent. The result being that Software Engineers can do a better job of reasoning about a design. This happens in much the same way that the Database Engineer will use the meta-data of an Entity Relationship Diagram to reason about the design of his or her database.

Please keep posting about your work.

We currently have a paper under reviewing for this year’s PLoP conference, where we show how to maintain the integrity of the model during a system evolution. Just stay tuned for more

Cynics think this is a result of the MDSD community desperately searching for a way to remain relevant, fans of MDSD regard it as a sign of MDSD’s superior sophistication.

I think we are definitely part of that second bunch

I’ve been changing my mind about DSLs lately. I think a programming language that would include a DSL for expressing domain models would have a lot of advantages.
The C# / LINQ combination is an example of something of the sort, where a general purpose programming language is extended with a querying language.
Adding such an extension to express domain models would be great. I don’t think plain classes are enough for this any more.

How can you ensure the integrity when you’re changing the rules on the fly? What about ongoing transactions on objects which are being modified in runtime? And context (sessions) becoming obsolete or invalid after some data rearranging?