Pierre-Louis Curien and Giorgio Ghelli. Decidability and Confluence of fijtop Reduction in F . Information and Computation, 1/2:57--114, 1994.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....with more interesting equivalences. Lillibridge [10] considered a language in which equivalence depends on the typing context. He eliminates the context sensitivityby tagging each path with its enclosing typing context, and then gives a rewriting strategy for this tagged system. Curien and Ghelli [5] gave a proof of decidability of term equivalence in F with fij reduction and a Top type. Because their Top type is both terminal and maximal, equivalence depends on both the typing context and the type at which terms are compared. They eliminate context sensitivityby inserting explicit coercions ....

Pierre-Louis Curien and Giorgio Ghelli. Decidability and Confluence of fijtop Reduction in F . Information and Computation, 1/2:57--114, 1994.

....with more interesting equivalences. Lillibridge [10] considered a language in which equivalence depends on the typing context. He eliminates the context sensitivity by tagging each path with its enclosing typing context, and then gives a rewriting strategy for this tagged system. Curien and Ghelli [5] gave a proof of decidability of term equivalence in F with fij reduction and a Top type. Because their Top type is both terminal and maximal, equivalence depends on both the typing context and the type at which terms are compared. They eliminate context sensitivity by inserting explicit coercions ....

Pierre-Louis Curien and Giorgio Ghelli. Decidability and Confluence of fijtop Reduction in F . Information and Computation, 1/2:57--114, 1994.

No context found.

P.-L. Curien and G. Ghelli. Decidability and confluence of fijtop reduction in F . Information and Computation, 109(1, 2):57--114, 1994.

....proved in [Ghe90] Although fij reduction in itself is not confluent, the confluence of reduction can be regained by adding a Top rule, which equates all terms with type Top. The fijTop system is normalizing (every term has a normal form) but it is still unknown whether it is terminating as well [CG94] Finally, it has been proved that the extension of system F with recursive types is not conservative, namely that, once recursive types are added, the traditional subtype checking algorithm is no longer complete, even with respect to non recursive types [Ghe93] On the other hand, there have ....

....not confluent in any calculus with a non trivial subtype relation. This problem has been addressed in system F by proving that fij reduction can be made confluent by adding a Top rule, which equates every term with a Top type, plus some rules which may be obtained by a Knuth Bendix like process [CG94] The same approach may also apply to system F bounded , but we leave this as an open issue. 4 Transitivity elimination Transitivity plays a central role in every subtype system. In fact, requiring the transitivity of the subtype relation is fundamental both from a conceptual and from a ....

[Article contains additional citation context not shown here]

P.-L. Curien and G. Ghelli. Decidability and confluence of fijtop reduction in F . Information and Computation, 109(1, 2):57--114, 1994.

No context found.

Pierre-Louis Curien and Giorgio Ghelli. Decidability and Confluence of fijtop Reduction in F . Information and Computation, 1/2:57--114, 1994.

No context found.

Pierre-Louis Curien and Giorgio Ghelli. Decidability and Confluence of fijtop Reduction in F . Information and Computation, 1/2:57--114, 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