There are many fascinating processes in the Universe, which we observe in more and more in detail thanks to increasing sophistication in technology. One of the most interesting phenomena is the life cycle of stars, their birth, evolution and death. If the stars are massive enough, they end their lives in the core-collapse supernova explosion, the most spectacular and violent event in the Universe. As the result, the most dense objects in the Universe, neutron stars and/or black holes are created. Naturally, we would like to understand the physical basis of these events and model and predict them in line with observation.

Taking the example from the life of stars, I will focus on a more general question whether the current status of our knowledge of processes in nature is adequate for their true understanding. I will show that we can construct many models, but so far their predictive power is limited or non-existent and, in most cases, we cannot tell right from wrong. Many models are dependent on current fashion in physics and, interestingly, on geography - Japan, Europe, Australia and the USA may have different believes. Finally a possible way forward to more progress in modelling of physical processes will be outlined, including the quark-meson-coupling model (QMC) which has a natural explanation for the saturation of nuclear forces and depends on very few adjustable parameters, strongly constrained by the underlying physics. Latest QMC results for compact objects and finite nuclei will be presented.