|
|
May, 2003 |
|
|
||
|
UML 2.0: The First Major Change to UML Since 1995
Thank You Karl Frank for our May 13 Meeting! Can it really be eight years since the introduction of UML at OOPSLA '95 and Jim Rumbaugh's well remembered presentation of the Unified Method to COOUG in December that year?  Karl Frank, a principal consultant at Borland, and a member of the OMG's
Analysis and Design Task Force, shared his insights on UML 2.0 at COOUG on
May 13. Much of the work on UML 2.0 focused on the underlying infrastructure
as a separate component for the UML.
Among the features which will be of interest to practitioners are the
emergence of aggregation/composition as a first-class modeling element,
along with the ability to represent aggregation or composition within a
package in new ways (in addition to the diamond symbol). One key aspect
which many will appreciate is that UML 2.0 will provide the
 underlying
infrastructure for exporting diagrams - as well as the underlying model
elements - from one modeling tool to another.
As practitioners, for the most part, we will have to "wait and see" what UML
2.0 will offer us; Because UML 2.0 componentizes the UML itself, different
vendors may respond to proposals for the different components. The major
work on the infrastructure will be of keen interest to tool vendors, but
there does not appear to be any major changes which would disrupt any UML
practices that are commonly used in our industry. Looking forward we would
hope that the metamodel refinements and interoperability standards would
lead to more and better tools.
Reported by Susan Burk and Terry McAuliffe |
