"JSF compliant" diagram for component diragam

* The new Cade book (JEE 5) doesn't give a solutioon using JSF.
* The old Cade book (purple cover, J2EE 1.4)) gave a solution using JSP + Servlets + EJBs.

I have a solution using JSP + Servlets + EJBs, for Part 2, but I want it to be using JSF.

To make it "JSF compliant" diagrams, what happens to the Servelts?

Prople have written here that a FacesServlet is provided for free with JSF.

OK, well how does this all translate to:

- Component Diagram (the detailed version; I plan to have two component diagrams)
- and Sequence Diagrams: Actor -----> JSPs -----> ??? ------> EJB ------>

A) "JSF compliant" solution leads to all custom Servlets removed? Only FacesServlet left?

B) "JSF compliant" solution leads to Managed Bean appearing in places? (As part of the design, which I am not changing, is that where I have session state, it is in a SFSB.)

B) "JSF compliant" solution can still have all places I have JSPs remaining as.... being called JSPs? (Another author suggested "JSF Page" stereotype -- a bit abstract?)

Many thanks.

As you noted there is just the built in servlet. You can indicate this for clarity. Or not. It's really up to you.

a) You wouldn't have any custom servlets.
B) Yes, the managed bean should show up as it is key to JSF.
B) I was fairly inconsistent and used a mix of JSP/JSF for the file names. I made a note in my submission with the logic behind why I did this.

I think as long as you document why you went with an approach and have a logical reason, it is ok.

I would like to make the following point.

As far as UML standard is concerned, which is supposed to be adhered to as per Sun's assignment (going by my personal experience), there are certain limitations imposed on your assuming things.

However, Sun has allowed their own guidelines to be violated, e.g. by using non-UML standards, as per experiences of participants of this forum. Hence, probably, people may get away with certain deviations from standards.

Still, the question remains, "Where lies the exact boundaries of such "pragmatic UML" [term coined by Jacobson] to be acceptable?"

Probably, you may have to chalk out your own strategy in absence of anything official, since no individual test-taker can vouchsafe whether her/his strategy actually worked, since s/he may not be cent-percent sure whether her/his assumptions held water with evaluators, even if the test taker succeeded.

P Das wrote:Probably, you may have to chalk out your own strategy in absence of anything official, since no individual test-taker can vouchsafe whether her/his strategy actually worked, since s/he may not be cent-percent sure whether her/his assumptions held water with evaluators, even if the test taker succeeded.

This is an excellent point. Since we don't get a score per diagram if we pass, we don't actually know what they liked/didn't.

Jeanne Boyarsky wrote:

B) I was fairly inconsistent and used a mix of JSP/JSF for the file names. I made a note in my submission with the logic behind why I did this.

Hi Jeanne,

I think in the class diagram we should use *.jsp or *.xhtml since we are showing just pages. The question is how JSFs should be shown in the Component Diagram? Would it be ok to show component with <<JSF>> stereotype and list all JSPs that are used by this component like in Cade & Sheil example?

I actually showed my JSPs in the class diagram with a << JSF >> stereotype - but I gave the extension .jsp and in the component diagram I did not give individual names...

and yes, not giving individual marks is a drawback for us - even if we had cleared the exam, we can't say for sure about individual points...

Do you think is a good idea to show <<JSF>> on the class diagram? I have read that ClsD should be web framework agnostic and Component Diagram is the one that shows which framework is used. In my opinion we should add <<JSF>> stereotype in the component diagram.

Hi guys,

I think my question quite relevat to this thread. Please share your idea.

Below are the my JSF modelling:

1. Class diagram
concrete JSF pages --> Controller --> concrete SLSBs

2. Sequence diagram:

concrete JSF pages --> Controller --> concrete Backing Beans --> concrete SLSBs

My assumption is that the Controller used here is Facelet.

Is everything ok with my modelling?

Dannyf Soon wrote:

2. Sequence diagram:

concrete JSF pages --> Controller --> concrete Backing Beans --> concrete SLSBs

I would not show controller on sequence diagram since if you are using JSF is alwasy FacetServlet - a framework class. You may show concrete backing beans since they are your classes.

One question, how do you show "concret JSF page" in the sequence diagram? Do you use any stereotype?

Krzysztof Koziol wrote:
I would not show controller on sequence diagram since if you are using JSF is alwasy FacetServlet - a framework class. You may show concrete backing beans since they are your classes.

One question, how do you show "concret JSF page" in the sequence diagram? Do you use any stereotype?

I'm not sure either. I wanted to use <<JSF Page>>, but Cad's book says "The class diagram remains web framework agnostic.

Ranchers please clarify this question. Thanks!

Yes it should. I was asking about sequence diagram.

Krzysztof Koziol wrote:Yes it should. I was asking about sequence diagram.

For JSF Pages, can sequence diagram use same stereotype as what is used in class diagram? I.e. both use <<JSP Page>>, can?

But Cad's book don't use any stereotype in the sequence diagram. What do you think?