In main.pdf some topological constraints are explored that are usually imposed on manifolds.
Why do those even matter for the definition of a manifold? Could we just 'relax' the definition of a manifold a bit? Such questions are addressed in the text.