Forbidden rectangles in compacta (Q441414): Difference between revisions

This paper is motivated by the following well-known problem due to van Douwen: does there exist a homogeneous compact space containing a cellular family of size greater than \(\mathfrak{c}\)? What might a homogeneous compactum with large cellularity ``look like''? The author states that one way to make this precise is to ask for examples of directed sets \((D,\leq)\) such that any homogeneous compactum \(X\) with a local base \(\mathcal{B}\) satisfying \((\mathcal{B},\supseteq) \cong (D,\leq)\) will satisfy some lower bound of \(c(X)\). In this interesting paper, the author proves among other things that that if a \(T_3\) space \(X\) has a neighborhood filter cofinally equivalent to a finite product \(\prod_{i\leq n} \kappa_i\) of regular cardinals \(\kappa_0 < \cdots < \kappa_n\), then \(c(X) \geq \kappa_n\). This he uses to conclude that in a homogeneous compactum, no neighborhood filter is cofinally equivalent to such \(\prod_{i\leq n} \kappa_i\). And also that that the supremum of the cardinalities of the free sequences in a homogeneous compactum \(X\) is always attained if GCH holds. Some interesting results on \(\beta\omega\) are also presented. For example, it is shown that the Fubini square and Fubini cube of a filter on \(\omega\) are always cofinally equivalent. (The ordering is containment.) This answers a question of Dobrinen and Todorcevic.