Rapide applications can be deployed to a wide variety of application servers or as standalone desktop applications.
The application is typically developed with separate client and server procedure steps, though a 'fat' client approach is also supported where the client and server logic is included in the same procedure step or load module.
For client/server applications, the server procedure steps are implemented as standard Gen servers on any Gen supported platform, for example CICS or IMS (COBOL), Windows or UNIX (C) and EJB (Java). This means that if you are converting an existing Gen application, there are no changes to your servers and they will continue to support the new Rapide client.
The client code is implemented using a combination of Gen generated Java code for the action diagram logic and Rapide generated code for the user interface and dialog flows.
 |
For browser based applications, the client logic is deployed on an application server and the user interface is displayed using a web browser.
|
 |
For desktop applications, the client logic is implemented as .jar files that execute locally on the desktop machine using a Java Virtual Machine (JVM). Once the application has been deployed to the desktop machine, the software can be easily and automatically updated without requiring any reboots or installation of third party runtimes, making maintenance of desktop applications much easier than in the past. |
The same source code is used to deploy the application to all platforms: browser and desktop, thus enabling easy deployment to multiple platforms and consistency in behaviour across all devices and platforms.
- The Rapide and Gen generated code can be generated from the Gen
toolset using a combination of the Gen Java generators and a Rapide
generator plug-in.
- Code can also be generated from the client/server encyclopaedia.
- The Rapide application is built and assembled using either the Rapide Generator plug-in or automated using GuardIEn.