Thanks, but note no claim was made of needing fsms to handle events [in component graphs]! /g I merely stated the [conjectured] motivation for UML's design team to introduce this level of sementics at the notational level.