We present a unifying framework for type systems for process calculi. The core of the system provides an accurate correspondence between essentially functional processes and linear logic proofs; fragments of this system correspond to previously known connections between proofs and processes. We show how the addition of extra logical axioms can widen the class of typeable processes in exchange for the loss of some computational properties like lock-freeness or termination, allowing us to see various well studied systems (like i/o types, linearity, control) as instances of a general pattern. This suggests unified methods for extending existing type systems with new features while staying in a well structured environment and constitutes a step towards the study of denotational semantics of processes using proof-theoretical methods.