In this paper we initiate further discussion on phase diagrams and structural transitions in tetrahedrally bonded semiconductors with particular emphasis on GaN. Comprehensive experimental data and the most important theoretical predictions are comprised in an effort to bring up a generalized universal scheme that explain melting that may or may not be accompanied by the change of the coordination number. The scheme considers the "anomalous" melting temperature dependence on pressure for GaN. Principal possibility of GaN melting into low density semiconducting liquid is demonstrated and challenges on the way to experimental observation of such liquid phase are disputed. Prospects of future experiments that can provide reliable information on the nature of atomic bonding for different GaN phases are laid out.