Designing affordances for the navigation of detail-on-demand hypervideo