\(\mathfrak{c}\)-many types of a \(\Psi\)-space (Q1755426)

Recall that a family \(\mathcal{A}\) of subsets of \(\omega\) is said to be almost disjoint if \(A \cap B\) is finite, for every \(A\) and \(B\) in \(\mathcal{A}\). Given an almost disjoint family \(\mathcal{A}\) on \(\omega\) one can define a topology on \(\Psi(\mathcal{A})= \omega \cup \mathcal{A}\) as follows: every point of \(\omega\) is isolated and a basic neighbourood of a point \(A \in \mathcal{A}\) is a set of the form \(\{A\} \cup A \setminus F\), where \(F\) is a finite subset of \(\omega\). This is known as the \(\Psi\)-space of \(\mathcal{A}\). The authors provide two constructions of continuum many almost disjoint families on \(\omega\) such that the associated \(\Psi\)-spaces are not homeomorphic. The first is a construction of branch families of \(2^\omega\). To every subset \(X\) of \(2^\omega\) we can associate an almost disjoint family of subsets of \(2^{<\omega}=\bigcup \{2^n: n < \omega \}\) in the following way: \(\mathcal{A}_X=\{\{x \upharpoonright n: n < \omega \}: x \in X\}\). Since \(2^{<\omega}\) is countable we can identify it with \(\omega\). Bernal-Santos and Garcia-Ferreira asked the following very natural question: suppose \(X\) and \(Y\) are two homeomorphic subsets of \(2^\omega\). Is it true that the spaces \(\Psi(\mathcal{A}_X)\) and \(\Psi(\mathcal{A}_Y)\) are homeomorphic? The authors give a strong negative answer to this question by constructing a family of continuum many homeomorphic subsets of \(2^\omega\) such that the \(\Psi\)-spaces of the associated almost disjoint families are not homeomorphic. In the second construction the authors provide continuum many Luzin families whose \(\Psi\)-spaces are not homeomorphic. An almost disjoint family \(\mathcal{A}\) of cardinality \(\omega_1\) is said to be Luzin if it has an enumeration \(\{A_\alpha: \alpha < \omega_1 \}\) such that for every \(\alpha < \omega_1\) and for every \(n < \omega\), the set \(\{\beta < \alpha: A_\beta \cap A_\alpha \subseteq n \}\) is finite.