We give a new definition of admissible representations which allows to handle also non countably-based topological spaces in the framework of Type-2 Theory of Effectivity. We prove that admissible representations δx, δy of topological spaces x,y have the desirable property that every partial function f :⊆ X → Y is continuously realizable with respect to δx, δy if and only if f is sequentially continuous. Furthermore, we characterize the class of the spaces having an admissible representation. Many interesting operators creating new topological spaces from old ones are shown to preserve the property of having an admissible representation. In particular, the class of sequential spaces with admissible representations turns out to be cartesian-closed. Thus, a reasonable computability theory is possible on important non countably-based spaces.
[1]
Viggo Stoltenberg-Hansen,et al.
Concrete Models of Computation for Topological Algebras
,
1999,
Theor. Comput. Sci..
[2]
Rüdiger Valk,et al.
Foundations of Computer Science: Potential - Theory - Cognition, to Wilfried Brauer on the occasion of his sixtieth birthday
,
1997
.
[3]
Marian Boykan Pour-El,et al.
Computability in analysis and physics
,
1989,
Perspectives in Mathematical Logic.
[4]
A. Grzegorczyk.
On the definitions of computable real continuous functions
,
1957
.
[5]
Christoph Kreitz,et al.
Theory of Representations
,
1985,
Theor. Comput. Sci..
[6]
Klaus Weihrauch.
A Foundation for Computable Analysis
,
1997,
SOFSEM.
[7]
Klaus Weihrauch,et al.
Computable Analysis: An Introduction
,
2014,
Texts in Theoretical Computer Science. An EATCS Series.