Tutorial
In developing runner, we use certain terms to describe the runner:
Database: The ASE database that stores individual atomistic simulation data, and further data used to define the run, as rows in the database.
Runner: A stateless spooling utility that reads from the database, and manages atomistic simulation workflows using a workflow manager.
RunnerData: The data attached to each row of database that instructs the runner on how to handle the row calculation.
Relay: A recorder to record the computational graph, needed to perform a workflow. It eases setting up complex workflows, with their runnerdata, and their submission and data collection.