Thinking over how to automate the time of data flows made me thing about data types. If the flows are going to be dynamic, then we need to embed metadata information within the flow. We can do this through the use of command codes. If the system sees a command code, it will modify itself (within limits) to be able to handle the new data type/size/speed/etc. This gives the system a LOT more potential, but makes it more complex and probably slower as well. We might even have to use a virtual machine to run it.