Last Modified: June 12, 2003
Note: It is highly recommended to have completed Tutorial
Part I before trying this example.
In this example, a Swing application showcases the Java Speech API
and
works as follows: the user chooses predefined phrases from a dropdown
box and the speech synthesizer pronounces them. As a motivation for
partitioning, imagine a scenario when this application needs to be run
on a machine, e.g. a PDA, that either has no speakers (hardware resource)
or does not have the Speech API installed (software resource). The idea
is to partition the original application in a client-server mode so that the
graphical partition (i.e. the GUI) would run on a PDA and it will control
the speech partition which would run on a desktop machine. We choose
this example because it fits well into the realm of applications amenable
for automatic application partitioning. The locality patterns here are very
clear and defined by the specific hardware resources (graphical screen
and speech synthesizer) and their corresponding classes (Swing and Speech
API).
In-order to run this example, you have to download Cloudgarden's
Java Speech API.
1) Once the Speech API package is successfully installed, download and
unjar Speech.jar
that contains all the file of the example.
2) Load Speech.xml configuration and perform the distribution.
(See Tutorial I for detailed
instructions).
Now you ready to run the partitioned version of the application:
3) Run the server at ./step4/desktop: startServer.bat or startServer.sh
4) Run the client part at ./step4/pda: java
remotecapable.speech.SpeechWindow
Last Modified:
June 12, 2003