Collaborative Learning through Flexible Web CVE - Teaching through Multi-User Virtual Environments - page 482

2. Event : is a trigger that can be raised by some
users' interactions or by the system itself,
notifying that something has happened in the
shared environment. Events can be grouped
into three main types, depending on the dif-
ferent kinds of interaction by which they are
caused:
a. Indirect active events : they are active
since it is the user himself who raises
the event, but indirect as they are not
generated by a direct voluntary user's
action, such as a mouse click on some
entity, but it is the system which raises
them, reacting for example to a particu-
lar user's position. A typical example
of such an event is OnProximity , raised
whenever an entity is closer to another
one than a fixed value.
b. Direct active events : in this case it is for
the user voluntarily to raise the event
interacting directly with an entity
c. Passive events : they are generated by
a shared system state change (i.e. posi-
tion update event for some entity).
3. Action : see paragraph on Virtual Actions.

of the system which listens to events propagated
within the virtual environment and reacting to
them performing one or more actions.
Action target can be indifferently the same
entity, which caused in some way the event, or
another different one.
The group made up of Event and Action takes
the name of Behavior as it states the manner in
which some entity evolves the system state when
perturbed.
Using this model we can design and code
different behavioral situations; for example the
following pseudo code shown in Algorithm 2,
which states at the same time:

• The way the object should be represent-

ed in the CVE, through the <Geometry>
element

• the way it reacts to events, in one word

how it behaves, through the element
<Behavior>

As the structured XML elements representing
the entities' appearance and behavior make use of
intuitive concepts and of a human readable form,
we can be confident that even non-programmers
without programming skills or any knowledge of
scripting languages can understand, manage and
produce a valid XML world instance that will
configure the runtime environment producing the
3D representation and the interactions expected.
Moreover the XML code and component reuse
reduces the time to create a particular experience
instance, abates the error rate and allows verifying
in real time every change made on the setup of
the virtual environment.

According to the classical view the Event
Conditioned Actions' rules are based on the fol-
lowing form:

On event
If conditions
Do actions

Moreover in our declarative model, the cen-
tral concepts around which all the scene graph is
described are the objects and users , or, in more
general worlds, entities .
Then we derived the following entity-oriented
interaction control model (Figure 5):
In this model the main component is an entity
that can be thought of as an abstraction of both
users, as active actors of the system allowed to
raise events and objects, as passive components

Collaborative Metaphors

Collaborative Metaphors (Barbieri & Paolini,
2001b) are a set of rules to support interaction and
collaboration between users who want to explore
complex content and information together. The
rules determines how the “collaborative com-