This article presents some reflections on how the recently published Boolean gene regulatory network (GRN) model for sea urchin endomesoderm development affects the problem of what we can expect to know about a developmental process. The Boolean computation demonstrated that, on a system-wide level, a topological GRN model can contain sufficient regulatory information to predict in silico all the spatial and almost all the temporal processes of regulatory gene expression observed in this phase of embryonic development. Conclusions that can be drawn illuminate the general and fundamental characteristics of developmental regulatory systems, such as their innate hierarchy and their reliance on logic-processing functions. The automaton-like performance which the Boolean model displayed reflects the basic quality of genomically controlled developmental process. This quality is of course the underlying requirement for a genetically encoded developmental mechanism. The accessibility of system-wide mechanistic explanation is something new in developmental biology, and turns on their head old truisms that for a century have been implicit in science aimed at small parts of systems.