The user interface for decisions and formulas

The IDIOM user interface makes extensive use of wizards to guide users in creating rules that accurately reflect business requirements and are free of construction errors. Drag and drop is used widely for its intuitive clarity and its ability to speed up the work of the adept user.

One of the rules builder’s first tasks when confronted with a new schema is to identify the various output fields (the places where the results of his/her formulas will be returned) and capture the sequence in which the input values are validated and used to generate the output values. This process flow can become quite sophisticated where there are dependencies between fields and output values that need to be built up by a succession of collaborating steps. To support this task IDIOM uses a mind-mapping metaphor in which graphical decision hierarchies are developed with increasingly detailed branches. In the following simple example, the leaf nodes in the hierarchy (rounded, blue icons) are decisions, which are executed at runtime in top to bottom order. The elements on the left of the Scope window (square, green icons) are decision groups whose function is to express the process decomposition.

Decision hierarchies show the decomposition of a decision request into the smallest logical units that



are useful to the business: that is, single atomic decisions. An atomic decision generates one output value: that is, one value that is returned for use by the business application (outside the rules engine itself). There are computed values that are not exported from the rules engine: these are ephemeral values calculated by the formulas that generate the atomic results. The rules engine is stateless and never retains such results.