Re: New version of AUI meta-model

Dear all,

since I still do not appear as a participant of the group, I have to fill
your mailbox with a (small) picture.

This is a fragment of exercise, promised last thursday, on the modeling of
an AUI for a (very basic) banking system.

The interface supports identification, selection of a transaction, and
execution of the transaction in a suitable container. The modeling is quite
generic, possibly as not generic as it could be, For example, one might
replace some selection with input or viceversa, but the idea would be that
selections are for pre-defined elements and input are more flexible, Anyway
one might also have a common abstraction fot that.

Also, one could actually refine the model by defining specific units for
the different supports For example, in this model some units are not
considered composed, but they might become so as the analysis proceeds, in
a refined model.

In general at this stage, we are defining that ultimately these units will
have to ensure that the indicated supports are provided.

Of course the model could be expanded with other types of transactions, and
this is just a starting point to give an idea. Each event support would
produce the corresponding event, giving rise to the corresponding behaviour.

In particular, the selection support would start a navigation behaviour to
activate the unit in which to perform the transaction, while the input
support would produce input events to feed data to some domain behaviour
activity. Of course, suitable presentations could be given as feedback, and
so on.

The model can represent a common abstraction for ATM interfaces, fragments
of an Internet Banking site, or even for a cashier terminal.

I hope this is useful

best
paolo

Received on Wednesday, 19 December 2012 13:33:56 UTC