323d53c8e939582e.tex
1: \begin{abstract}
2: There are several ideal boundaries and completions in General
3: Relativity %(g-boundary, b-boundary, etc.)
4:  sharing the topological
5: property of being  {\em sequential}, i.e.,  determined by the
6: convergence of its sequences and, so,  by some limit operator $L$. As emphasized in a classical article by Geroch, Liang and Wald, some of them have the property,
7: commonly regarded as a  drawback, that there are points of the
8: spacetime $M$ non $T_1$-separated from points of the
9: boundary $\partial M$.
10: 
11: Here we show that this problem  can be solved
12: from a general topological viewpoint. In particular, there is a canonical minimum refinement of the topology in the completion $\overline{M}$
13: which T$_2$-separates the spacetime $M$ and its boundary $\partial M$  ---no matter the type of completion one chooses.
14: Moreover,
15: %we focus on
16:  % the process to define a topology in terms of a limit operator $L$ (which characterizes the possible convergence of its sequences)
17:   %, as commonly used in relativistic completions. This procedure does not make sense for any topological space but it does in the very general class of
18:  we analyze the case  of sequential  spaces and  show how the refined $T_2$-separating topology
19:  %for the completion
20:  can be constructed from a modification $L^*$ of the original limit operator $L$. Finally, we particularize this procedure to the case of the causal boundary and show how the separability of $M$ and  $\partial M$ can be introduced as an abstract axiom in its definition.
21: 
22: \end{abstract}
23: