First try worked already. Probing the interaction between substrate curvature and instabilities in thin films. On a flat plane, a hole in a fluid collapses symmetrically, but on cylinder the front is unstable. What about the intermediate case?

Long story short: extracting power law exponents from even very clean numerical data is hard, but there are some good tricks around :). Do check the erratum that fixes a small, strange but possibly very confusing typo that crept in…

How to mix very viscous fluids? Add some particles while stirring. Ice cubes may work but simple glass beads also work just fine. Thanks Vittorio for contributing.

Not quite accurate enough for a plate, but the solvent trap works.