Skip to content
Leon Starr edited this page Jan 23, 2022 · 3 revisions

R53 / 1:Mc

Process Input Port must consume exactly one Process Input Source

Process Input Source feeds zero, one or many Process Input Port


Think about the computation defined inside of a Process (by the client). By requiring that all input ports be connected, we ensure that no special logic is required to handle cases where input is not available. This is especially important in a data flow graph since each Process requires all of its inputs to be available in order to execute.

That said, it is perfectly okay to send the empty set to any Process Input Port. In that case we can mark the data as being available and enable execution.

If data from multiple sources must be merged somehow, an intermediate Process must be defined to handle the merge into a single output. There is no implicit merging of data on a Process Input Port.

Formalization

Process Input Port.(Input source, Graph) -> Process Input Source.(ID, Graph)
Clone this wiki locally