Sequence diagrams' details

Hi,
a couple of questions regarding SCEA pt 2.:

1) If I interpret the emphasis on visual-feedback as a permission to use AJAX - should I put the AJAX calls on the sequence diagrams? How much detail for this "secondary" AJAX functionality would be expected?

2) What's the set of objects that we should consider on sequence diagrams (how detailed should they be)? Should there be entity managers there (vide Cade/Sheil) ? Proxies for external system as well as the external system itself? Caches?

Thanks in advance and regards

Pedro,
Welcome to CodeRanch!

1) Yes. If the AJAX functionality is part of your use case, it needs to be in a sequence diagram. As an example, support my use case involves doing a query, seeing the results and sorting. I would have three "entry points" to my sequence diagram for that use case. If all of this is done in AJAX, I still have three entry points. It just so happens that two of them are AJAX and one is a full page load.

2) There isn't going to be a right answer to this. As much detail as is necessary to solve the problem. I ended with the call to entity manager on mine.

Showing the Entity Manager in the Sequence diagram, does that mean that we need to show it in the Class diagram as well ?

I added 'EntityManager' in the class diagram with a stereotype '<< JPA >>'. However, I did not add any methods (though method calls are shown in the Sequence Diagram).