In this paper it is shown how structural and cognitive versioning issues can be efficiently managed in a Petri nets based hypertextual model. The advantages of this formalism are enhanced by a modular and structured modeling; modularity allows to focus the attention only on some modules, while giving the abstraction of the others. Each module owns meta knowledge that is useful in defining new layers and con texts.
The central point of the data model is the formulation and resolution of three recurrence equations, effective in describing both the versioning and the derivation history; these equations permit to express in precise terms both the structural evolution (changes operated on specific nodes of the net) and the behavioral one (changes concerning browsing).
