Type transformations with complicated transition conditions

[joeseibel]

When type transformations are used to determine the error type of a transition's target state, what should be the rules when a transition has a condition that is a complicated expression. In section E.8.3, paragraph (9), the standard says, "In the case of multiple contributors, e.g., a conjunction in a transition trigger condition, the transformation rule is applied repeatedly in the order of the conjunction elements." However, this sentence doesn't really explain the rules for determining the target state's error type.

What should be the rules for or, and, all, ormore, and orless expressions? What about nested and repeated expressions? Should transformations even apply to transitions with expressions or should they only apply to transitions that simply reference a single event or propagation as the transition condition?

[brlarson]

Seems like restriction of transitions to single events is the only coherent resolution.