Reputation: 567
The pictures below shows a sequences of messages They represent the following situations :
My question is : as long as these respresent the logic inside the case description it is OK right ? Or does UML imposes some sort of logic to follow for the way message (stimuli) and their responses (return) mustbe sequenced ?
Pic 1
Pic 2
EDITED PICTURES
Pic 1
Pic 2
Upvotes: 1
Views: 85
Reputation: 36305
Well, I think that both have flaws since you use asynchronous messages for the self-messages. While that might be valid, it's most likely a wrong use here. Send an asynchronous message to self means what? A message in a bottle that will be found later? Some fork-message? Now assuming you meant a synchronous message, you probably want to extent the life line from above since the Check Name Length will likely be a sequential action somewhen after the user name is received.
wrt edit pict 1 If that is you intention, it is fine. However, I once again doubt that. It is called Check Name Length, and unless you are doing that for pure fun, you will likely somewhere deal with the result of the check. And if this is asynchronous, then how will you know if the check is complete?
Upvotes: 1