| < Day Day Up > |
|
A distributed program is one that runs on multiple computers. These types of programs are concurrent because parts of them are run on physically different CPUs; therefore, these systems can often be modeled as components and can use the techniques described in the text. Distributed programming has evolved from very simple mechanisms using technologies such as Unit-to-UNIX Copy (UUCP) and sockets into techniques that take advantage of the higher-level languages in which the programs are written. Perhaps the best example of this is Java remote method invocation (RMI), which uses the Java interfaces and exceptions to create distributed programs that almost completely hide the complexity of a distributed program.
This chapter presents distributed programming using RMI. It first gives a simple example of a "HelloWorld" program that covers the basics of RMI, how to set up a simple RMI program, and how to actually make the program run. The chapter then shows how objects can be created to run on the computer where they were created or to be physically transferred to a remote computer and run on the remote system. This behavior is then combined with the Java Event Model to create a chat program, such as the UNIX "talk" command or AOL Instant Messenger.
| < Day Day Up > |
|