Hyland, M. and Ong, L., Pi-calculus, dialogue games and PCF, FPCA'95, ACM, 1995.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....to establish full abstraction. While we expect a direct behavioural proof would be possible, the correspondence, in addition to facilitating the proof, o ers deeper understanding of the present type discipline and game semantics. We brie y give comparisons with related work. Hyland and Ong [26] presented a calculus encoding of innocent strategies of their games and show operational correspondence with a calculus encoding of PCF. Fiore and Honda [12] propose another calculus encoding for call by value games [22] Our work, while being built on these preceding studies, is novel in ....

....z) m) M ] m j m(z )z [ i ( u ) P i j Arghu x1 : xn 1 zi ) where P0 = N ] u 0 else P i = L] u 0 . x : M : u = m) u m] j [ M : m j [x m] Arghx yzi [ Nat] def = x( y ) i [y i j [z z] Nat Fig. 2. Encoding of PCF [26], are given in Figure 2. Copy cat processes are given by [x x x[ i ( y i ) x 0 in i ( and, for replicated types, x x ] Copy cats for unary types are special cases where the indexing sets are singletons. The interpretation of [ 1 : n Nat] says a process, ....

Hyland, M. and Ong, L., Pi-calculus, dialogue games and PCF, FPCA'95, ACM, 1995.

....axiomatic frameworks for polymorphism with recursion (accommodating both domain theoretic models [CGW89] and per models [AP90, FMRS90] together with a corresponding representation theory. Game semantics. The application of game theory to semantics has provided new insights and results [AMR94, HO95b, AM95] In particular, AMR94] and [HO95a] have constructed intensionally fully abstract semantic models of PCF which yield fully abstract models by extensional collapse. As advocated by Abramsky, axiomatic studies of game semantics may lead to abstract full abstraction results (as had happened ....

J.M.E. Hyland and C.-H.L. Ong. Pi-calculus, dialogue games and PCF. In 7th Annual ACM Conference on Functional Programming Languages and Computer Architecture, La Jolla, California, 1995.

.... Shortly after Plotkin s paper, Milner proved that there was exactly one fully abstract model of PCF meeting certain conditions [16] Until recently, all descriptions of this fully abstract model have used operational semantics (see, for instance, 19, 34] New constructions using games semantics [2, 10, 20] and logical relations [22, 33] have yielded a more abstract understanding of PCF. The sequentiality problem is enduring because it is robust. For instance, changing the reduction strategy of PCF from call by name to call by value makes no difference: versions of the parallel or function ....

....a case expression can branch) Until recently, it was not known how to adopt games models to the call byvalue setting. Honda and Yoshida have bridged that gap, devising a model for call by value PCF using games semantics [9] The model loosens the restrictions of the original games semantics [2, 10, 20] to include strategies that start with the opponent s answer rather than a question. Intuitively, this means that the value supplied to a call by value function is immediately available without interrogation by the player. The basic definitions are quite different from our logical relations based ....

J. M. E. Hyland and C.-H. L. Ong. Pi-calculus, dialogue games and PCF. In 7th Annual ACM Conference on Functional Programming Languages and Computer Architecture, La Jolla, California, 1995.

....only a weaker notion of behavioural properties than the original ones: the essential content of types is partially lost through the translation. Gaining insights from the preceding studies on types for interaction including types for the calculus [33, 39, 17, 48, 31] and game semantics [3, 4, 28, 29, 25], the present authors, with Martin Berger, recently reported [6, 49] that basic type structures for the calculus which precisely capture existing type structure do exist, allowing fully abstract translation of prominent functional typed calculi. In [6, 49] we have presented two type disciplines ....

.... is given in Appendix A, the output x( y)Q acts as ( y) xh yijQ) in the notation with free name passing) Bound name passing has an essentially equivalent expressive power as free name passing [42, 18] and is convenient to obtain precise correspondence with functional type structures [6, 28, 49]. We assume the branching allows for any countable indexing set with cardinality more than one. Branching is used for representing base values as well as conditionals [6] and plays an essential r ole in the information flow analysis later. The reduction relation is generated by the following ....

Hyland, M. and Ong, L., Pi-calculus, dialogue games and PCF, FPCA'93, ACM, 1995.

....it is worth acknowledging the work that others have undertaken on the basic concepts of games. Abramsky, Jagadeesan, and Malacaria [AJP94] are the originators of the specific form of games described in this report. Other types of games have also been proposed by Blass [Bla92] and Hyland and Ong [HO94, HO95]. Games naturally incorporate a notion of linearity and there is a strong correspondence with Girard s linear logic [Gir87] The syntax chosen for the calculus associated with linear logic used in this report has been particularly influenced by the work of Wadler [Wad93] and Bierman [BBdPH92] ....

J. M. E. Hyland and C.-H. L. Ong. Pi-calculus, dialogue games and PCF. In Proceedings of the 7th acm Conference on Functional Programming Languages and Computer Architecture, pages 96--107. ACM Press, 1995.

....states, events, and resources [15, 28, 29] A connection between linear logic and randomized interaction was investigated in [22] which introduced a game semantics of a fragment of linear logic by means of probabilistic games. Other notions of game semantics for linear logic were considered in [8, 1, 2, 17, 20, 19, 11]. Another direction is emphasized here: linear logic proof game simulations of probabilistic games from complexity theory. Such simulations are then used to show that certain problems in linear logic that are known to be hard to decide are also hard to approximate. We shall consider so called ....

J. M. E. Hyland and C. H. L. Ong, Pi-calculus, dialogue games and PCF, Proceedings of the 7th ACM Conference on Functional Programming Languages and Computer Architecture, ACM Press, 1995, pp. 96--107.

....to establish full abstraction. While we expect a direct behavioural proof would be possible, the correspondence, in addition to facilitating the proof, o ers deeper understanding of the present type discipline and game semantics. We brie y give comparisons with related work. Hyland and Ong [23] presented a calculus encoding of innocent strategies of their games, and show operational correspondence with a calculus encoding of PCF. Fiore and Honda [10] propose another calculus encoding for call by value games [19] Our work, while being built on these preceding studies, is novel in ....

.... ( n ) Nat) Now the syntax of PCF terms are given by: M : x j x : M j MN j n j succ(M) j pred(M) j ifzero M then N else L j x : M We omit operational semantics and the typing rules [13] The mappings from PCF types and terms to types and terms, which are due to Hyland and Ong [23], are given in Figure 2 (copy cat [x y] is de ned in Appendix B) The 14 interpretation of [ 1 : n 1 Nat] says a process, when asked for its value, asks back questions at types 1 ; n 1 , receives the results to these questions, and nally returns a natural number as the answer to the ....

Hyland, M. and Ong, L., Pi-calculus, dialogue games and PCF, FPCA'95, ACM, 1995.

....states, events, and resources [15, 28, 29] A connection between linear logic and randomized interaction was investigated in [22] which introduced a game semantics of a fragment of linear logic by means of probabilistic games. Other notions of game semantics for linear logic were considered in [8, 1, 2, 17, 20, 19, 11]. Another direction is emphasized here: linear logic proof game simulations of probabilistic games from complexity theory. Such simulations are then used to show that certain problems in linear logic that are known to be hard to decide are also hard to approximate. We shall consider so called ....

J. M. E. Hyland and C.-H. L. Ong. Pi-calculus, dialogue games and PCF. In Proceedings of the 7th acm Conference on Functional Programming Languages and Computer Architecture, pages 96--107. ACM Press, 1995.

.... Linear logic, introduced in [10] is a refinement of classical logic often described as being resource sensitive because of its intrinsic ability to reflect computational states, events, and resources [11, 27, 28] Several notions of game semantics for linear logic are investigated in [6, 1, 2, 13, 17, 15, 9]. Connections between linear logic proof search and probabilistic games considered in complexity theory are investigated in [20, 21, 22] In particular, linear logic proof search may also be seen as a game. This game, the linear logic proof game, is played on linear logic formulas, and its moves ....

J. M. E. Hyland and C.-H. L. Ong. Pi-calculus, dialogue games and PCF. In Proceedings of the 7th acm Conference on Functional Programming Languages and Computer Architecture, pages 96--107. ACM Press, 1995.

....exposes all basic issues. Mutable references only di er in that setref destructively updates the reference. For the process encoding, immutable and mutable references are quite di erent, as given in Figures 6 and 7, respectively. We use the so called copy cat agents coming from games semantics [5, 19, 18]. There are a few suggestions from the encoding. A couple of observations: 1) Low level references may as well refer to high level values and vice versa, following our secrecy annotation in types: this makes sense practically (for example when one wishes to pass receive a low level object which ....

Hyland, M. and Ong, L., Pi-calculus, dialogue games and PCF, FPCA'93, ACM, 1995.

.... programme of synthetic domain theory, where one tries to integrate the different approaches by working in, for example, the effective topos [7] One should also 2 remark that intensional aspects may nonetheless play a role in the study of equational full abstraction see the study of games in [1, 8]. The rest of the paper is structured as follows. In Section 2 we review the definitions of the three versions of PCF that we will consider. We also define the syntax of a program logic for these languages, and propose a simple operational interpretation of this logic. In Section 3 we introduce a ....

....to construct a simulator for the relevant language within itself. A very similar method has recently (and independently) been employed by Abramsky et al. 1] to prove a definability result for an interpretation of sequential PCF based on games; a similar result has been proved by Hyland and Ong [8]. This yields an alternate semantic proof of the existence of enumerators for PCF, analogous to that in [14] for PCF . It is worth remarking that there is no corresponding definability result for PCF . It may well be that there can be none; it is not at all clear, however, how to even ....

[Article contains additional citation context not shown here]

J. M. E. Hyland and C.-H. L. Ong, Pi-calculus, Dialogue Games and PCF, in Proc. 7th ACM Conf. Functional Programming and Computer Architecture, ACM Press, 1995; see also On Full Abstraction for PCF: I, II and III, by the same authors, to appear.

....dependency in terms of communicated values into consideration. The above informal principle based on causal dependency 1 is simple, but may look basic as a way of stipulating information flow for processes. Since many language constructs are known to be representable as interacting processes [6, 24, 25], one may wonder whether the above idea can be used for understanding safety in information flow in var 1 Related ideas are studied in the context of CCS [12] and CSP [39] 3 ious programming languages. In the following, we consider this question by taking basic examples of information flow ....

.... type # i ) As one notable aspect, we only use those processes whose outputs (in any of three forms) are bound, e.g. each unary output has a form x(# #y : # #) P (this restricted output is an important mode of communication which arises in the context of both # calculus [37] and games semantics [6, 25, 24]) Accordingly we set names in each vector instantiating agent variables to be pairwise distinct. These restrictions make typing rules simpler, elucidate basic nature of the present notion of types, and give enough descriptive power to serve our present purpose. 3) The secrecy index s guarantees ....

[Article contains additional citation context not shown here]

Hyland, M. and Ong, L., Pi-calculus, dialogue games and PCF, FPCA'93, ACM, 1995.

.... power, uniformly representing diverse language constructs as name passing processes, including those of imperative, functional and object oriented languages, as well as sequential and concurrent, cf. 39, 54] These representations are in close relationship with those given in games semantics, cf. [31, 30, 13], which suggest precise type structures in which the corresponding representation in name passing processes can be articulated. Since many real life programming languages are equipped with diverse constructs from di erent paradigms, it would be interesting to see whether we can obtain a typed ....

....based on causal dependency among interaction 1 is simple, 1 Related ideas are studied in the context of CCS [14] and CSP [47] 4 but may look basic as a way of stipulating information ow for processes. Since many language constructs are known to be representable as interacting processes [6, 7, 30, 31, 38, 39], one may wonder whether the above idea can be used for understanding safety in information ow in various programming languages. In the following, we consider this question by taking basic examples of information ow in imperative programs. 2.2. Syntax Let a; b; c; x; y; z; range ....

[Article contains additional citation context not shown here]

Hyland, M. and Ong, L., Pi-calculus, dialogue games and PCF, FPCA'93, ACM, 1995.

....dependency in terms of communicated values into consideration. The above informal principle based on causal dependency 1 is simple, but may look basic as a way of stipulating information ow for processes. Since many language constructs are known to be representable as interacting processes [6, 24, 25], one may wonder whether the above idea can be used for understanding safety in information ow in various programming languages. In the following, we consider this question by taking basic examples of information ow in imperative programs. 1 Related ideas are studied in the context of CCS [12] ....

.... a type i ) As one notable aspect, we only use those processes whose outputs (in any of three forms) are bound, e.g. each unary output has a form x( y : P (this restricted output is an important mode of communication which arises in the context of both calculus [38] and games semantics [6, 25, 24]) Accordingly we set names in each vector instantiating agent variables to be pairwise distinct. These restrictions make typing rules simpler, elucidate basic nature of the present notion of types, and give enough descriptive power to serve our present purpose. 3) The secrecy index s guarantees ....

[Article contains additional citation context not shown here]

Hyland, M. and Ong, L., Pi-calculus, dialogue games and PCF, FPCA'93, ACM, 1995.

.... power, uniformly representing diverse language constructs as name passing processes, including those of imperative, functional and object oriented languages, as well as sequential and concurrent, cf. 39, 54] These representations are in close relationship with those given in games semantics, cf. [31, 30, 13], which suggest precise type structures in which the corresponding representation in name passing processes can be articulated. Since many real life programming languages are equipped with diverse constructs from di#erent paradigms, it would be interesting to see whether we can obtain a typed ....

....based on causal dependency among interaction 1 is simple, 1 Related ideas are studied in the context of CCS [14] and CSP [47] 4 but may look basic as a way of stipulating information flow for processes. Since many language constructs are known to be representable as interacting processes [6, 7, 30, 31, 38, 39], one may wonder whether the above idea can be used for understanding safety in information flow in various programming languages. In the following, we consider this question by taking basic examples of information flow in imperative programs. 2.2. Syntax Let a, b, c, x, y, z, range ....

[Article contains additional citation context not shown here]

Hyland, M. and Ong, L., Pi-calculus, dialogue games and PCF, FPCA'93, ACM, 1995.

....be localized, and this leads to chronicles and agendas. By the way we are not quite producing a coherent space, since D;E might be neither evenly or oddly coherent. The notions defined below are presumably related to the views of A H 2 O games B introduced by Hanno Nickau [7] and Hyland Ong [5]. 18 Jean Yves Girard Definition 17 Let D = m 0 ; m p ] be a dispute, and let b be a bloke in m i 1 which is not in m i (there is at most one such i) then b has a well defined father in m i , namely a bloke a of the same parity (b is said to be a son of a) with the notation of ....

J. M. E. Hyland and C.-H. L. Ong. \Pi-calculus, dialogue games and PCF. In Proc. 7th acm Conf. Functional Prog. Lang. Comp. Architecture, pages 96 -- 107. acm Press, 1995.

.... based on computable functions will be naturally too restrictive (note also that it is essentially impossible to semantically embed these calculi in each other [47] suggesting how subtle semantic difference in functional computations made explicit when they are represented as processes [17, 24]) This situation is similar to the case when we consider compositional embedding between these two functional calculi, cf. 47] however, computational behaviour in calculus is based on much fine grained name passing nondeterministic interactions than functional one. The question then ....

Hyland, M. and Ong, L., Pi-calculus, dialogue games and PCF, FPCA'93, ACM, 1995.

....essence, parallelism in a language with state imposes a choice between nondeterminism or efficiency. The naive models thus prevent one from deriving principles about determinate stateful programs surely an important class of programs. Recent descriptions of sequentiality using game semantics [2, 10, 15] and logical relations [23, 16] have achieved precise models of PCF, with different advantages and disadvantages. Game semantics explains sequentiality via a dialogue between a term and its surrounding program; this explanation may tie in well with various theories of concurrent processes. The ....

J. Hyland and C.-H. L. Ong. Pi-calculus, dialogue games and PCF. In 7th Annual ACM Conference on Functional Programming Languages and Computer Architecture, La Jolla, California, 1995.

.... [8, 34] By concentrating on specific forms of interaction which obey a few basic constraints, the approach makes it possible to extract desired classes of interacting processes at a high level of abstraction, offering suitable semantic universes for varied calculi and programming languages, cf. [2, 3, 4, 19, 20, 24]. The forms of interaction in these universes are however inherently call by name: it has LFCS, Department of Computer Science, University of Edinburgh. e mail: kohei dcs.ed.ac.uk, ny dcs.ed.ac.uk. Supported in part by EPSRC Fellowships and JSPS Research Fellowships. not been clear how the ....

Hyland, M. and Ong, L., Pi-calculus, dialogue games and PCF, FPCA'93, ACM, 1995.

.... which should obey a few basic constraints, the approach makes it possible to extract desired classes of interacting processes at the high level of abstraction, offering suitable semantic universes for varied calculi and programming languages including those with imperative features, cf. [2, 3, 4, 5, 26, 31]. The forms of interaction in these universes are however inherently call by name: it has not been clear how the call by value computation can be captured in the setting of game semantics, in spite of its equally significant status as a mode of computation. LFCS, Department of Computer Science, ....

....proofs of the basic properties of the operational structures of call by value Hyland Ong games (the proofs can often be applicable to call by name) The appendices also include a brief introduction to the calulus representation of call by value strategies. 2. Games and Strategies Hyland and Ong [26] introduced a notion of arenas and innocent strategies for their full abstraction result for PCF. Strategies are a class of interacting processes representing the interactive aspects of programs behaviours, while arenas give a basic notion of types for these processes. In [26] these notions are ....

[Article contains additional citation context not shown here]

Hyland, M. and Ong, L., Pi-calculus, dialogue games and PCF, Proceedings of 7th Annual ACM Conference on Functional Programming Languages and Computer Architecture, June 26-28, 1995.

....Science, Centre of the Danish National Research Foundation. one fully abstract model of PCF meeting certain conditions [16] Until recently, all descriptions of this fully abstract model have used operational semantics (see, for instance, 19, 34] New constructions using games semantics [2, 10, 20] and logical relations [22, 33] have yielded a more abstract understanding of PCF. The sequentiality problem is enduring because it is robust. For instance, changing the reduction strategy of PCF from call by name to call by value makes no difference: versions of the parallel or function ....

....a case expression can branch) Until recently, it was not known how to adopt games models to the call by value setting. Honda and Yoshida have bridged that gap, devising a model for call by value PCF using games semantics [9] The model loosens the restrictions of the original games semantics [2, 10, 20] to include strategies that start with the opponent s answer rather than a question. Intuitively, this means that the value supplied to a call by value function is immediately available without interrogation by the player. The basic definitions are quite different from our logical relations based ....

J. M. E. Hyland and C.-H. L. Ong. Pi-calculus, dialogue games and PCF. In 7th Annual ACM Conference on Functional Programming Languages and Computer Architecture, La Jolla, California, 1995.

....introduction in [27] many works have shown the stunning expressive power of this calculus to represent diverse classes of computational behaviours, in spite of its simple communication primitive, the name passing. As examples, mutable and immutable data structures [27, 26] higher order functions [24, 18], parallel objectoriented languages [35] agent passing calculi [32] mobile protocols [30] as well as semantic universes with intensional structures [18, 15] are all known to be representable as a combination of name passing interaction. In these representations, original high level operations ....

....its simple communication primitive, the name passing. As examples, mutable and immutable data structures [27, 26] higher order functions [24, 18] parallel objectoriented languages [35] agent passing calculi [32] mobile protocols [30] as well as semantic universes with intensional structures [18, 15] are all known to be representable as a combination of name passing interaction. In these representations, original high level operations are decomposed into a series of name passing communications, while preserving the compositional meaning of the original dynamics. Thus the universe of ....

Hyland, M. and Ong, L., Pi-calculus, dialogue games and PCF, FPCA'95, ACM Press, 1995.

.... which should obey a few basic constraints, the approach makes it possible to extract desired classes of interacting processes at the high level of abstraction, offering suitable semantic universes for varied calculi and programming languages including those with imperative features, cf. [2, 4, 5, 6, 28, 33]. The forms of interaction in these universes are however inherently call by name: it has not been clear how the call by value computation can be captured in the setting of game semantics, in spite of its equally significant status as a mode of computation. y) LFCS, Department of Computer ....

....thank anonymous referees for their beneficial suggestions. We are grateful to Samson Abramsky, Vincent Danos, Guy McCusker, Pasquale Malacaria, Paul Mellies and Jon Riecke for their comments and or discussions, and to N. Raja for his hospitality in Bombay. 2. Games and Strategies Hyland and Ong [28] introduced a notion of arenas and innocent strategies for their full abstraction result for PCF. Strategies are a class of interacting processes representing the interactive aspects of programs behaviours, while arenas give a basic notion of types for these processes. In [28] these notions are ....

[Article contains additional citation context not shown here]

Hyland, M. and Ong, L., Pi-calculus, dialogue games and PCF, Proceedings of 7th Annual ACM Conference on Functional Programming Languages and Computer Architecture, June 26-28, 1995.

.... by introducing a category of games G i with optimised strategies; we show that the denotational semantics in G i gives a compilation of FPC terms into core Pict codes (the asynchronous polyadic calculus without summation) The process representation follows a pioneering idea of Hyland and Ong [18]. However, we advance their representation by introducing semantically well founded optimisation techniques; we also extend the setting to encompass the rich type structure of FPC. The resulting code gives basic insight on the relationship between the abstract, categorical, types and their ....

....resulting code gives basic insight on the relationship between the abstract, categorical, types and their possible implementations. 1 Introduction Games in semantics. Recently, the notions of games and strategies have been used for constructing mathematical models of programming languages; e.g. [4, 1, 17, 7, 18, 2, 22, 15, 3]. The basic common idea underlying these works is to construct a universe of semantic domains in which a program phrase is modelled as an object with internal structure reflecting its computational behaviour in an abstract way. In particular, semantic points are interacting Research supported ....

[Article contains additional citation context not shown here]

M. Hyland and L. Ong. Pi-calculus, dialogue games and PCF. In FPCA '95 Conf., pages 96--107, 1995.

....omit. 5.4. Embedding: Types. It is well known that the choice of evaluation strategies affects the embedding of calculi in calculus, including their sorting [21, 23] We here choose the call by value evaluation, since this gives a more straightforward encoding in the presence of constants (cf. [13]) The embedding for the call by name dynamics is briefly discussed later. In the following, we translate each ; type into a a sorting considered as a tree (with nodes labelled by sorts and data sets, indexed edges induced by the sorting function) which we call sorting tree. These trees are ....

....) where [a = b] ff is recursively given as: 1) a(xy) x 0 y 0 )bhx 0 y 0 i: x 0 = x] ff 1 j[y 0 = y] ff 2 ) if ff = ff 1 ff 2 s.t. ff 1 62 D) 2) a(xy) y 0 )bhxy 0 i: y 0 = y] ff 2 (ff = d ff 2 ) and (3) a(x) bhxi (ff 2 D) this is a copy cat behaviour, cf. [2, 13, 28]) Given appropriate notions of convergence, we have M : ff iff hh M : ffii u (cf. 21] Let [ E M : ff] wu def = w(x 1 : x n ) M ] u , with E = x 1 : ff 1 ; x n : ff n . Then hhE; ffii wu hhE M : ffii wu is a sorted term. We now set [ E M : ff] A wu as [ hhE; ffii wu ....

[Article contains additional citation context not shown here]

Hyland, M. and Ong, L., Pi-calculus, dialogue games and PCF, FPCA'95, ACM Press, 1995.

....axiomatic frameworks for polymorphism with recursion (accommodating both domaintheoretic models [CGW89] and per models [AP90, FMRS90] together with a corresponding representation theory. Game semantics. The application of game theory to semantics has provided new insights and results [AMJ94, HO95b, AM95] In particular, AMJ94] and [HO95a] have constructed intensionally fully abstract semantic models of PCF which yield fully abstract models by extensional collapse. As advocated by Abramsky, axiomatic studies of game semantics may lead to abstract full abstraction results (as had happened ....

J.M.E. Hyland and C.-H.L. Ong. Pi-calculus, dialogue games and PCF. In 7th Annual ACM Conference on Functional Programming Languages and Computer Architecture, La Jolla, California, 1995.

.... Linear logic, introduced in [12] is a refinement of classical logic often described as being resource sensitive because of its intrinsic ability to reflect computational states, events, and resources [13, 32, 33, 23] Several notions of game semantics for linear logic are investigated in [6, 1, 2, 15, 19, 17, 9]. Connections between linear logic and probabilistic games considered in complexity theory are investigated in [24, 27, 25] In particular, linear logic proof search may also be seen as a game. This game, the linear logic proof game, is played on linear logic formulas, and its moves are instances ....

J. M. E. Hyland and C.-H. L. Ong. Pi-calculus, dialogue games and PCF. In Proceedings of the 7th acm Conference on Functional Programming Languages and Computer Architecture, pages 96--107. ACM Press, 1995.

....introduction in [74] many works have shown the stunning expressive power of this calculus to represent diverse classes of computational behaviours, in spite of its simple communication primitive, the name passing. As examples, mutable and immutable data structures [74, 71] higher order functions [69, 56], parallel object oriented languages [99] agent passing calculi [92] mobile protocols [86] as well as semantic universes with intensional structures [56, 53] are all known to be representable as a combination of name passing interaction. In these representations, original high level operations ....

....its simple communication primitive, the name passing. As examples, mutable and immutable data structures [74, 71] higher order functions [69, 56] parallel object oriented languages [99] agent passing calculi [92] mobile protocols [86] as well as semantic universes with intensional structures [56, 53] are all known to be representable as a combination of name passing interaction. In these representations, original high level operations are decomposed into a series of name passing communications, while preserving the compositional meaning of the original dynamics, so that diverse computational ....

[Article contains additional citation context not shown here]

Hyland, M. and Ong, L., Pi-calculus, dialogue games and PCF, FPCA'95, ACM, 1995.

....linear logic [21] has also renewed interest in game theory and functional programming languages. The refined connectives of linear logic have helped shed new light on work on games [12] The games models have proved useful for programming language semantics: the recent fully abstract models of pcf [2, 26] are good examples of this. In addition, intuitionistic linear logic (ILL) has been proposed as a resource sensitive foundation of functional programming languages. Thus it would seem useful to reconsider the work on CL in a linear setting, viz. to reconsider classical linear logic (CLL) ....

J.M.E. Hyland and C.-H.L. Ong. Pi-calculus, dialogue games and PCF. In Proceedings of Conference on Functional Programming Languages and Computer Architecture, pages 96 -- 107, 1995.

....and answers in which one player may use scratchpads for book keeping purposes. The model is a concrete instance of category and it satisfies a completeness property. Innocent approach: a review The basic setting has much in common with the innocent approach of Hyland and Ong (see [18] or [19] for an extended abstract) a brief account of which follows. A basic arena, a, is given by a question a (there is no harm in overloading the symbol a) and a set Ans(a) of possible answers. It is useful to bear in mind the following reading: Interpretation a Ans(a) logical atomic proposition ....

J. M. E. Hyland and C.-H. L. Ong. Pi-calculus, dialogue games and PCF. In Proc. 7th acm Conf. Functional Prog. Lang. Comp. Architecture, pages 96 -- 107. acm Press, 1995.

No context found.

Hyland, M. and Ong, L., Pi-calculus, dialogue games and PCF, FPCA'95, ACM, 1995.

No context found.

J. M. E. Hyland and C.-H. L. Ong, Pi-calculus, Dialogue Games and PCF, in Proc. 7th ACM Conf. Functional Programming and Computer Architecture, ACM Press, 1995; see also On Full Abstraction for PCF: I, II and III, by the same authors, to appear.

No context found.

Hyland, M. and Ong, L., Pi-calculus, dialogue games and PCF, FPCA'95, ACM, 1995.

No context found.

Hyland, M. and Ong, L., Pi-calculus, dialogue games and PCF, FPCA'95, ACM, 1995.

No context found.

J. M. E. Hyland and C.-H. L. Ong. Pi-calculus, dialogue games and PCF. In 7th Annual ACM Conference on Functional Programming Languages and Computer Architecture, La Jolla, California, 1995.

No context found.

Hyland, M. and Ong, L., Pi-calculus, dialogue games and PCF, FPCA'93, ACM, 1995.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC