The good news is I have found out today that I passed stage I, result wasn't too great (80%)

but I'm pleased with the overall result. Now for the hard part !. Please can you guys give me some advice on how to go about stage II. Specifically, what I would like to know is, is there a correct order in comprising the UML components ? By this I mean, based on what is supplied in the assignment ( Use Cases etc) :-
1) Does one first comprise the Class Diagrams ?, i.e are the Class Diagrams the starting point for the assignment submission ?
2) Do sequence diagrams begin after Class Diagrams (Since you have to list Classes at the top ?) or do you begin with sequence diagrams (and thus deducing what classes you require for your class diagram ??)
3) To construct the class diagrams - is this a process of extending the classes given in the business domain model and supplementing that with classes identified from the detailed USE CASES supplied (i.e looking for nouns etc in the USE CASE requirements ?)
4) Is it necessary to model two separate clients ? i.e a WEB CLIENT and a SWING CLIENT ? They both make and initiate requests although the protocol may be different (HTTP/RMI) - would it be plausible to identify my client as as UI CLIENT and in my assumptions state that this can either be WEB or SWING based ?
5) Is there a UML notation to model constraints ? i.e lets say I was building an ATM banking system and I wanted to ensure that latency after a person entered their pin code was under 2 seconds - Can this be modelled ?
Thanks everyone !!
Priya Patel