Data stream model helps developers to build "chains" of connected computing nodes and run them. A computing node has input/ouput sockets, and the "dsm runner" is responsible to carry "data streams" through the "chains" and feed "data streams" to computing nodes.
Usage:
Please read the main function in Composite.py for an example.
Notes:
Developers are responsible to
- build computing nodes (with i/o sockets)
Developers or Users are responsible to
- specify connections of computing nodes
Hints for developers:
- Read tests in source codes to understand how to use dsm. Start with Composite.py and Runner.py is a good idea
Limitations:
- No loops are allowed.
- Currently only a trivial implementation of "runner" exists. So data streams are limited in one physical machine.