Abstract not found

. We recast dataflow in a modern categorical light using profunctors as a generalisation of relations. The well known causal anomalies associated with relational semantics of indeterminate dataflow are avoided, but still we preserve much of the intuitions of a relational model. The development fits with the view of categories of models for concurrency and the general treatment of bisimulation they provide. In particular it fits with the recent categorical formulation of feedback using traced monoidal categories. The payoffs are: (1) explicit relations to existing models and semantics, especially the usual axioms of monotone IO automata are read off from the definition of profunctors, (2) a new definition of bisimulation for dataflow, the proof of the congruence of which benefits from the preservation properties associated with open maps and (3) a treatment of higherorder dataflow as a biproduct, essentially by following the geometry of interaction programme. 1 Introduction A fundament...

This paper addresses the problem of defining a formal tool to compare the expressive power of different concurrent constraint languages. We refine the notion of embedding by adding some "reasonable" conditions, suitable for concurrent frameworks. The new notion, called modular embedding, is used to define a preorder among these languages, representing different degrees of expressiveness. We show that this preorder is not trivial (i.e. it does not collapse into one equivalence class) by proving that Flat CP cannot be embedded into Flat GHC, and that Flat GHC cannot be embedded into a language without communication primitives in the guards, while the converses hold. 4 A; C; D; G; M;O;P;R; T : In calligraphic style. ss; ff ; dd: In slanted style. \Sigma; \Gamma; #; oe; ; /; ø; ff. S ; [; "; ;; 2 j=; 6j=; ; 9 +; k; ~ +; ~ k; ! \Gamma! W ; \Gamma! ; ; \Gamma! W ; \Gamma! ; h; i; [[; ]]; d; e ffi; ?; ; 5 All reasonable programming languages are equivalent, since they are Turing...

Given suitable categories T; C and functor F : T ! C, if X; Y are objects of T, then we define an (X; Y )-relation in C to be a triple (R; r; ¯ r), where R is an object of C and r : R ! FX and ¯ r : R ! FY are morphisms of C. We define an algebra of relations in C, including operations of "relabeling," "sequential composition," "parallel composition," and "feedback," which correspond intuitively to ways in which processes can be composed into networks. Each of these operations is defined in terms of composition and limits in C, and we observe that any operations defined in this way are preserved under the mapping from relations in C to relations in C 0 induced by a continuous functor G : C ! C 0 . To apply the theory, we define a category Auto of concurrent automata, and we give an operational semantics of dataflow-like networks of processes with indeterminate behaviors, in which a network is modeled as a relation in Auto. We then define a category EvDom of "event domains," a (non...

We analyze the relative expressive power of variants of the indeterminate fair merge operator in the context of static dataflow. We establish that there are three different, provably inequivalent, forms of unbounded indeterminacy. In particular, we show that the well-known fair merge primitive cannot be expressed with just unbounded indeterminacy. Our proofs are based on a simple trace semantics and on identifying properties of the behaviors of networks that are invariant under network composition. The properties we consider in this paper are all generalizations of monotonicity. 1 Introduction The study of indeterminate computing systems is motivated by seeking to understand the theory underlying what is commonly called "distributed systems" programming. In such systems indeterminacy is introduced when one abstracts away from low level hardware or timing details. The subject receives further im- Present Address: School of Computer Science, McGill University, Montreal, PQ, CANADA y...

We recast dataflow in a modern categorical light using profunctors as a generalization of relations. The well known causal anomalies associated with relational semantics of indeterminate dataflow are avoided, but still we preserve much of the intuitions of a relational model. The development fits with the view of categories of models for concurrency and the general treatment of bisimulation they provide. In particular it fits with the recent categorical formulation of feedback using traced monoidal categories. The payoffs are: (1) explicit relations to existing models and semantics, especially the usual axioms of monotone IO automata are read off from the definition of profunctors, (2) a new definition of bisimulation for dataflow, the proof of the congruence of which benefits from the preservation properties associated with open maps and (3) a treatment of higher-order dataflow as a biproduct, essentially by following the geometry of interaction programme.

A deterministic message-communicating process can be characterized by a "continuous" function f which describes the relationship between the inputs and the outputs of the process. The operational behavior of a network of deterministic processes can be deduced from the least fixpoint of a function g, where g is obtained from the functions that characterize the component processes of the network. We show in this paper that a nondeterministic process can be characterized by a "description" consisting of a pair of functions. The behavior of a network consisting of such processes can be obtained from the "smooth" solutions of the descriptions characterizing its component processes. The notion of smooth solution is a generalization of least fixpoint. Descriptions enjoy the crucial property that a variable may be replaced by its definition.

Kahn's principle states that if each process in a dataflow network computes a continuous input/output function, then so does the entire network. Moreover, in that case the function computed by the network is the least fixed point of a continuous functional determined by the structure of the network and the functions computed by the individual processes. Previous attempts to generalize this principle in a straightforward way to "indeterminate" networks, in which processes need not compute functions, have been either too complex or have failed to give results consistent with operational semantics. In this paper, we give a simple, direct generalization of Kahn's fixed-point principle to a large class of indeterminate dataflow networks, and we prove that results obtained by the generalized principle are in agreement with a natural operational semantics. 1 Introduction Dataflow networks are a parallel programming paradigm in which a collection of concurrently and asynchronously executing s...

Abstract We define a concrete operational model of concurrent systems, called trace automata. For such automata, there is a natural notion of permutation equivalence of computation sequences, which holds between two computation sequences precisely when they represent two interleaved views of the "same concurrent computation. " Alternatively, permutation equivalence can be characterized in terms of a residual operation on transitions of the automaton, and many interesting properties of concurrent computations can be expressed with the help of this operation. In particular, concurrent computations, ordered by "prefix, " form a Scott domain whose structure we characterize up to isomorphism. By axiomatizing the properties of the residual operation, we obtain a more abstract formulation of automata, which we call concurrent transition systems (CTS's). By exploiting a correspondence between concurrent alphabets and certain CTS's, we are able to use the rich algebraic structure of CTS's to obtain results in trace theory. Finally, we connect CTS's and trace automata by obtaining a characterization of those CTS's that correspond in a natural way to trace automata, and we show how the correspondence suggests an interesting notion of morphism of trace automata.

An automaton with concurrency relations A is a labeled transition system with a collection of binary relations indicating when two actions in a given state of the automaton can occur independently of each other. The concurrency relations induce a natural equivalence relation for finite computation sequences. We investigate two graph-theoretic representations of the equivalence classes of computation sequences and obtain that under suitable assumptions on A they are isomorphic. Furthermore, the graphs are shown to carry a monoid operation reflecting precisely the composition of computations. This generalizes fundamental graph-theoretical representation results due to Mazurkiewicz in trace theory.