M. Abadi. Baby Modula--3 and a Theory of Objects. Journal of Functional Programming, 4(2):249--283, 1994.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....and derivations, and then checked for soundness. Thus, in principle, labels can be made transparent from outside pro types (hence, to the user of the system) and only seen as internal devices needed to ensure type soundness. A system that exhibits features comparable to ours is Baby Modula 3 [Aba94] which, however, we generalize in two respects: i) we allow object extensions and subsumptions in any order, while in [Aba94] all the extensions must be done before any subsumption; ii) our completions may be extended as a result of a method addition, while in [Aba94] completions are fixed ....

....to the user of the system) and only seen as internal devices needed to ensure type soundness. A system that exhibits features comparable to ours is Baby Modula 3 [Aba94] which, however, we generalize in two respects: i) we allow object extensions and subsumptions in any order, while in [Aba94] all the extensions must be done before any subsumption; ii) our completions may be extended as a result of a method addition, while in [Aba94] completions are fixed ahead of time, prior to any addition. A feature of [Aba94] that, instead, we do not provide (even though we could) is the ....

[Article contains additional citation context not shown here]

M. Abadi. Baby Modula--3 and a Theory of Objects. Journal of Functional Programming, 4(2):249--283, 1994.

.... 6 considers an interesting encoding of classes as objects that share a lot of similarities with the object oriented language Smalltalk 80 [15] The last section is devoted to the comparison with the Lambda Calculus of Objects of [12] and related papers [14, 10, 21, 9, 18, 8, 7] Baby Modula 3 of [1], and contains also open problems and the conclusions. We assume that the reader is familiar with some object oriented concepts such as delegation based object calculi, type and subtype systems, self types. Some knowledge of the seminal papers [12, 5] and the above cited related papers) would be ....

....accept as input both an instance of the Point and C Point classes, and as a consequence the following judgments are derivable: f Point:new : int, and f C Point:new : int. 7 Related Work Among the many object based languages we find in the literature, we recall the following ones. M. Abadi, in [1], presents a small functional language which include the main features of Modula 3. This language allows object override, a small form of object extension and width subtyping. The soundness of the typing system is guaranteed by a denotational semantics. The Lambda Calculus of Objects of [12] is ....

M. Abadi. Baby Modula--3 and a Theory of Objects. Journal of Functional Programming, 4(2):249--283, 1994.

....sur les objets, semantique operationnelle, systeme de types. 1 Introduction In the last few years, the problem of designing safe and expressive type systems for object based languages (also called prototype based languages) has been widely addressed. The seminal works of [US87, CU89, Mic90, Aba94, FHM94, AC96a] share the same object oriented philosophy, where the main entity is the one of object instead of the one of class. In those papers, classes can be easily codified by appropriate objects, following the classes as objects analogy of Smalltalk 80 [GR83] In objectbased languages, ....

M. Abadi. Baby Modula--3 and a Theory of Objects. Journal of Functional Programming, 4(2):249--283, 1994.

....users to download code into the kernel. An operating system faces many challenges common to those facing ANs. The language chosen by the spin designers is Modula 3 which is a strongly typed language with its automatic garbage collection, and known denotational semantics for a Modula 3 subset [31]. The type checking of the Modula 3 compiler is used to ensure domains of protection at the level of binding names for access to objects. One of the seminal works in the field of ANs 18 was the thesis of Christian Tschudin [32] in which he used a form of Pascal as the basis for developing the ....

Martin Abadi. Baby modula-3 and a theory of objects. Journal of functional programming, 4(2), April 1994.

....of Harper s [19] using closures and stacks instead of formal substitutions. This approach yields a manageable proof for a realistic implementation strategy. A few other object formalisms have been defined and studied. Many of these rely on purely functional models, with an emphasis on types [1, 7, 11, 13, 20, 24, 25, 26, 32]. Others deal with imperative features in the context of concurrency; see for example [34] The works most closely related to ours are that of Eifrig et al. on LOOP [18] and that of Bruce et al. on PolyTOIL [9] LOOP and PolyTOIL are typed, imperative, objectoriented languages with procedures, ....

Abadi, M., Baby Modula-3 and a theory of objects. Journal of Functional Programming 4(2), 249-283. 1994.

....has furthered the theoretical development of programming languages and strengthened the connection between formal analysis of the object oriented features of a language and the development of a high level conceptual model of the language. Abadi s analysis of the core features of Modula 3, in [Aba93] has served as a model for the analysis we have applied to C . In addition, van Gent s thesis, vG93] and Schuett s thesis, Sch94b] have set very high standards for the depth of semantic analysis appropriate to an undergraduate thesis. 1.11 A Guide to this Thesis This thesis is designed to ....

Martin Abadi. Baby modula-3 and a theory of objects, April 1993.

....derivations, and then checked for soundness. Thus, in principle, labels can be made transparent from outside class types (hence, to the user of the system) and only seen as internal devices needed to ensure type soundness. A system that exhibits features comparable to ours is Baby Modula 3 [1] which, however, we generalize in two respects: i) we allow object extensions and subsumptions in any order, while in [1] all the extensions must be done before any subsumption; ii) our completions may be extended as a result of a method addition, while in [1] completions are fixed ahead of ....

....(hence, to the user of the system) and only seen as internal devices needed to ensure type soundness. A system that exhibits features comparable to ours is Baby Modula 3 [1] which, however, we generalize in two respects: i) we allow object extensions and subsumptions in any order, while in [1] all the extensions must be done before any subsumption; ii) our completions may be extended as a result of a method addition, while in [1] completions are fixed ahead of time, prior to any addition. A feature of [1] that, instead, we do not provide (even though we could) is the distinction ....

[Article contains additional citation context not shown here]

M. Abadi. Baby Modula--3 and a Theory of Objects. Journal of Functional Programming, 4(2):249--283, 1994.

....ECOOP (European Conference on Object Oriented Programming) 48 and OOPSLA (Object Oriented Programming: Systems, Languages, and Applications) conferences. The Journal of Function Programming published a special issue in 1994 on the theory of objectoriented languages which contains the papers [Aba94, Bru94, PT94] while the journal Theory and Practice of Object oriented Systems (TAPOS) will publish a special issue on types in object oriented languages in late 1996. Acknowledgements: Thanks to Luca Cardelli, Peter Wegner, Tony Simons and the anonymous referees for very helpful comments and ....

Martin Abadi. Baby Modula-3 and a theory of objects. Journal of functional programming, 4:249--283, 1994.

....users to download code into the kernel. An operating system faces many challenges common to those facing ANs. The language chosen by the spin designers is Modula 3 which is a strongly typed language with its automatic garbage collection, and known denotational semantics for a Modula 3 subset [32]. The type checking of the Modula 3 compiler is used to ensure domains of protection at the level of binding names for access to objects. 9 Future work ANs are an exciting possibility, but they also present a tremendous danger to a valuable shared resource. The first step in developing secure ANs ....

Martin Abadi. Baby modula-3 and a theory of objects. Journal of functional programming, 4(2), April 1994.

....of fixed points in determining the meaning of object oriented constructs. Another more recent paper (Pierce and Turner, 1992b) provides a general technique for modeling binary operations. This can be applied either to the approach of (Pierce and Turner, 1993) or to that given here. Recent work by Martin Abadi (1994) has resulted in a somewhat different style of semantics for the object oriented features of Modula 3. As Modula 3 is based on a delegation style of inheritance rather than the class based approach of the languages cited here, it is not clear whether this work can be extended to class based ....

Martin Abadi. Baby modula-3 and a theory of objects. Journal of functional programming, to appear, 1994.

....second order bounded quantification, while we require higher order bounded quantification. On the other hand, ours is simpler in that it omits recursive types in favor of existential types (which can themselves be encoded using only universal quantification) 30 B. C. Pierce and D. N. Turner Abadi (1993) and Mitchell, Honsell, and Fisher (1993) present related models of objects and delegation based inheritance (Ungar Smith, 1987) In both of these systems, a basic calculus like formalism is extended with new syntactic forms designed to directly capture the operations of message sending and ....

Abadi, M. 1993 (February). Baby Modula-3 and a Theory of Objects. Research Report 95.

....imperative features, with corresponding proof techniques. In further work [5] we treat second order type structures (with Self types) for an imperative calculus. A few other object formalisms have been defined and studied. Many of these rely on purely functional models, with an emphasis on types [1, 8, 10, 12, 17, 19 21, 26]. Others deal with imperative features in the context of concurrency; see for example [28] The works most closely related to ours are that of Eifrig et al. on LOOP [15] and that of Bruce and van Gent on TOIL [9] LOOP and TOIL are typed, imperative, object oriented languages with procedures, ....

Abadi, M., Baby Modula-3 and a theory of objects. Journal of Functional Programming 4(2), 249-283. 1994.

....is an important operation that it is characterized by peculiar typing rules; we prefer to study it directly and not to explain it away at an early stage. Similarly, the choice of objects with a fixed number of components, instead of extensible ones [Wand 1987; Mitchell, Honsell, Fisher 1993; Abadi 1994] is a conscious one. Without ruling out future work on extensible objects, we feel that fixed size objects are easier to handle, particularly in the later stages of our type theoretical development. Finally, we do not provide an operation to extract a method from an object as a function. As we ....

....methods [Smyth, Plotkin 1982] instead of metric methods. However, even the fragment F : poses problems, as noted in [Abadi, Plotkin 1990] We expect object types to give rise to further, interesting difficulties. Many of the ideas in our semantics originate in the study of Baby Modula 3 [Abadi 1994]. See section 11.1 for a comparison. 9.1 The Untyped Universe We interpret al..l our languages in an untyped universe. This section described the structure of this untyped universe. The assumptions on the universe are fairly standard and technical [MacQueen, Plotkin, Sethi 1986; Amadio 1989; ....

[Article contains additional citation context not shown here]

M. Abadi, Baby Modula-3 and a theory of objects. Journal of Functional Programming 4(2).

....feel that override is an important operation that June 5, 1995 11:11 AM Page 3 it is characterized by peculiar typing rules; we prefer to study it directly and not to explain it away at an early stage. Similarly, the choice of objects with a fixed number of components, instead of extensible ones [1, 15, 16, 21], is a conscious one. Without ruling out future work on extensible objects, we feel that fixed size objects are easier to handle, particularly in the later stages of our type theoretical development. Finally, we do not provide an operation to extract a method from an object as a function. As we ....

Abadi, M., Baby Modula-3 and a theory of objects. Journal of Functional Programming 4(2), 249-283. 1994.

No context found.

Abadi, M., Baby Modula-3 and a theory of objects. Journal of Functional Programming 4(2), 249-283. 1994.

....an encoding that does not respect subtyping. Pierce and Turner [Pierce, Turner 1994] sketched an encoding for objects and classes that respects subtyping, but does not account for our method override operations on objects. Some ideas presented here originated in the study of Baby Modula 3 [Abadi 1994]. That calculus resembles in power Ob 1 : Ob 1 : plus recursion) but the two are incomparable. The semantics of Baby Modula 3 is based on ideals. A per semantics is also briefly sketched, without a corresponding study of equational rules. Our work is closely related in spirit to that of ....

M. Abadi, Baby Modula-3 and a theory of objects. Journal of Functional Programming 4(2).

....imperative features, with corresponding proof techniques. In further work [5] we treat second order type structures (with Self types) for an imperative calculus. A few other object formalisms have been defined and studied. Many of these rely on purely functional models, with an emphasis on types [1, 8, 10, 12, 17, 19 21, 27]. Others deal with imperative features in the context of concurrency; see for example [29] The works most closely related to ours are that of Eifrig et al. on LOOP [15] and that of Bruce and van Gent on TOIL [9] LOOP and TOIL are typed, imperative, object oriented languages with procedures, ....

Abadi, M., Baby Modula-3 and a theory of objects. Journal of Functional Programming 4(2), 249-283. 1994.

No context found.

M. Abadi. Baby Modula--3 and a Theory of Objects. Journal of Functional Programming, 4(2):249--283, 1994.

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