AKA OMT. Developed by JimRumbaugh and others. The notation of OMT and much of its philosophy lie behind a lot of the UnifiedModelingLanguage and (to a lesser extent) the RationalUnifiedProcess.

OMT was the leading OOA/D method for a long time because it emphasised a structural approach to modelling that fitted well with the EntityRelationalModelling techniques that were already well known, and so provided an easy transition to OO for many organizations. 

At one time OMT also included "object-oriented DataflowDiagram". Ouch

One of the big ideas in OMT is that OO development is ''seamless''. The idea is that you model the problem in classes at some notional ''high level'', and then uniformly add more and more ''detail'' until eventually you get some code. 

This idea seems to be one root of the RoundTripEngineering myth. And the more I think about it there is strong smell of ScientificManagement about it, too.

CategoryMethodology