Collaborative Learning through Flexible Web CVE - Teaching through Multi-User Virtual Environments - page 477

in their virtual dome, different clues hid-
den in the virtual world, different questions
to answer.

evolve during the collaborative session through a
declarative description of the Event Conditioned
Actions (ECA) that can be performed in the virtual
environment.

On the grounds of the previous considerations,
the main goals of WebTalk04 ( WT04 ) architecture
were:

The Runtime environment

XML-Based Implementation

• To provide a flexible and highly decoupled

development environment to allow differ-
ent developer teams to design and deploy
specific 3D collaborative environments
dealing with only the specific task (3D
design, contents assembling, story board
design, software develop) they were expert
in, without regarding any other task.

In this section we will show how the WebTalk04
declarative system is implemented using a markup
language coded with XML Schema (XML-Sche-
ma: www.w3.org/XML/Schema).
Since a homogeneous encoding is provided
on all levels, this is a consistent approach for
all abstraction stages starting from scene graph
level up the description of the complex user's
interactions.
Moreover an XML representation is well
suited to describe the 3D scene as structured data
that can be processed without paying attention to
how the data should be presented. The conversion
stylesheets in fact allow the switch from one for-
mat to another. Through different XSL files, the
content of a WT04 XML scene graph file could
be easily converted to VRML or X3D, to pretty-
printed HTML, or to any number of other formats.
Expressions of scenes in XML enables appli-
cation of a wide range of existing and emerging
XML-based tools for transformation, translation
and processing. XML provides numerous benefits
for extensibility and componentization, as well as
the ability to develop well-formed and validated
scene graph, an extremely valuable constraint
since “broken” 3D content would no longer be
allowed to escape onto the Web where if might
cause larger scenes to fail.
Finally, a formal XML description of the
scene graph and the user-to-user or user-to-object
interactions seems the best “interface” between
the following subsystems:
WT04 runtime environment , in which this XML
file works as a declarative description of how the

• To allow it to easily define and set up a ses-

sion's prototype, meant as a set of linked
3D stages that works as a mere place where
the action takes place and sharing the same
kind of virtual experience (be it education-
al, entertaining, informative), main col-
laboration features and the same environ-
mental containers, from which are derived
different, context-specific instances of the
virtual world, well suited for different us-
ers and situations.

• To convey to these virtual environments

a high sense of virtual presence, leaving
the full graphic driven paradigm to switch
towards a new one that takes the overall
application design into account and to pro-
vide a set of interaction features to drive
the collaboration among users in the spe-
cific way the designer was intended to do.

We proposed, then, a declarative format to
reduce the programming need for a non-expert
designer of the collaborative experience. For
this reason we will examine the major issues
concerned in the formal description of a CVE
both from a static point of view, regarding how
the virtual word is when it is generated and from
a dynamic point of view regarding on how it can