Partitioning the projective plane and the dunce hat (Q2674561)

The partitionability of a simplicial complex is an important combinatorial property. For instance, shellable simplicial complexes are partitionable. The long-standing open Stanley's conjecture, now disproved, predicted that every Cohen-Macaulay simplicial complex is partitionable. Now, it is still open for two-dimensional Cohen-Macaulay simplicial complexes. This is the last open case since partitionability for \(1\)-dimensional simplicial complexes (i.e. graphs) is well-known. For a partitionable pure simplicial complex, the \(h\)-vector has a combinatorial meaning (see pg. 3). In this article, the author proves that any triangulation of the real projective plane is partitionable (and the same for any triangulation of the dunce hat, see Theorem 1). This result is extremely interesting for three reasons: the first is that, since partitionability is a combinatorial property, proving that every triangulation of a given topological space is partitionable provides deep insight into the spaces themselves. Moreover, it is significant evident to believe that Stanley's conjecture could be true in dimension two. Last, but not least, the technical results proved to get the main theorem are original and of independent interest, with various possible applications. In particular, the idea is to split the complex which we want to prove to be partitionable into partitionable subcomplexes and then glue them back to the original complex, preserving partitionability (see Lemma 3.1 and Lemma 3.4). Key intermediate results are given in Section 4. The final steps to prove Theorem 1 are given in Theorem 5.3 (for the real projective plane) and in Theorem 5.11 (for the dunce hat). The paper adopts the relative simplicial complexes' general point of view, proving also other interesting results (see, for instance, Lemma 3.8 and Lemma 3.11). The overall impression of the paper is very good because it is clear and well-written.