I have a question regarding UML Sequence diagram.
Let's say we have an object Customer, KioskUI, KioskService and the KioskService has the last process called showNotification(); does the method showNotification goes to the Customer object or KioskUI?
If KioskUI, does the object have to return a variable to the Customer object as "<----- notification", or we no longer have to return something from the KioskUI to the Customer?
Since sequence diagram only focus on how process operates with one another my other question is, is a tangible object like a printed report is no longer needed to be shown in the sequence diagram or does it have to?
In object oriented systems it is good to see sequence diagrams as depictions of objects sending messages to each to communicate. Arrows in these diagrams show the sender/receiver relation between objects in time, where the label for the arrow depict the message itself, e.g. a method call with arguments. You are asking if the backwards arrow should point to KioskUI or the Customer object. The answer is simple, it depends on who sent the message, e.g. where will the method call "showNotification()" return. You have to decide this, but I guess KioskUI is a natural choice. So the answer is, you should show the backwards arrow to the Customer only if there was an forward arrow from Customer (this should be depicted using the box on the lifeline unless it was an asynchronous call). What is needed to be depicted in a diagram is always a choice depending on your need and UML makes no difference as long as it is a Classifier I guess, which means anything which can have an instance can be there - classes, components, actors, nodes,... For example I saw many use case realizations done as sequence diagrams with actors as objects with lifelines.