Because IVE limits the type of data stored in its internal
database it is easier to develop new visualizations for IVE.
Each visualization is developed as a plug-in that has
access to the IVE internal storage and it can communicate
with other plug-ins through simple message dispatching
system, see Fig. 6. Visualization is instantiated using IVE
wizard that allows selection of visualization plug-in and
selection of data sets (log file, application model,
audio/video recording) that will be used in the
visualization. Each visualization plug-in can announce
which data sets are mandatory and it can also limit amount
of data sets. Then each visualization plug-in opens one or
more views that are placed into the main IVE window
using NetBeans built-in layout mechanism. Some
visualizations may wish to produce changes in other
visualizations, e.g. highlighting of the same information in
other visualization views. Instead of storing this
information in the internal database we introduced central
message dispatcher that collects and resends all messages
between all visualization plug-ins.