The underlying process execution engine, channels, connections, groups and workers are all independent of the process client. You can replace the process client with a different folder type and the client should remain usable through a basic user interface. The only things you have to do are:
- Implement or inherit bindings to locate the different parts of the process client.
- Include the Root path field to maintain the root path, which is used by some functionality such as sending emails.
This means that you could:
- Create and run processes as part of another application, but using the same mechanisms as used by the process client. At the simplest, you can use the Worker to run standalone processes without any use of groups, connections or anything else.
- Create an alternative to the process client, optimised for a particular use. For example, you could replace the process client with a more "push button" interface for use in a kiosk environment.
- Use the underlying channel method to integrate between applications, or to integrate directly into a process client.
These development are out of scope for this guide.