UML Behavioral Diagrams: State Transition Diagram - Georgia Tech - Software Development Process

preview_player
Показать описание
Рекомендации по теме
Комментарии
Автор

A real world example would be really nice

demisx
Автор

It was uncleared for me why diagrams offer actions either on transition or inside states. This video helped me a little to understand. Tell me if I'm wrong, the Yakindu site says it comes from historically Mealy and Moore state machines (1955): the 1st executes actions from its transitions, while the second from its states, but these two state machines describe exactly the same system without losing any expressiveness. In 1984, Harel improves these state machines and name it statechart: statecharts = state-diagrams + depth + orthogonality + broadcast communication. But why having allowed mixing the both Mealy/Moore notations ? I guess it's just to give more flexibility to the architect and to offer him to spread information all over the diagram, like this video is saying. I paraphrase it: actions from transitions are made in first, then actions from state "on entry" then "do" then "on exit". What this video does not explain explicitly is the state "event", when called, will disable "on entry" actions and "on exit" actions. The second point not well explained is when the state has a transition leaving and entering it, the "on entry" and "on exit" actions are not called (because already entered and left). Finally, the state "event" is to avoid drawing the transition on the state. All these are simplfy tips and tricks: multiple ways to tell the same things (kind of sugar syntax). All of this is poorly documented on the web. Now I have a better idea of all this works, I hope.

quentinquadrat
Автор

Not useful. Feels like the video was made on the deadline

manchul