Using DFT with up to hundreds of atoms per slab unit cell, I use several sources to illustrate basic principles concerning metal-oxide surfaces, interactions with adspecies and metal overlayers, and ultrathin (<1nm) oxide films on metals and adsorption thereon: 1) Metal oxides vary greatly in relaxation to adsorbates, with those of alumina among the largest. This surface becomes so locally oxidizing that multiply charged metal adatoms can occur. 2) The interface between an oxide film and a basal-plane metal substrate can differ fundamentally. Most metals prefer O(1x1), but the nobel metals prefer cations, which convert to a somewhat positive metallic layer with O-ions above. 3) Point defects can nucleate metal islands and alter growth and adhesion. However, common Fs and Vs centers destabilize metal dimers (the first step in nucleation), while another common defect, ad-OH, does the opposite. In fact, 1/3 ML of ad-OH on sapphire reverses the growth mode and more than doubles the adhesion of Cu metal. These results are simply due to geometry and electrostatics. 4) Ultrathin film adsorbate response depends on the substrate: While alumina on Ru is rather inert to a Pd overlayer, the film on Al(111) reacts strongly, drawing substrate metal into the oxide and extruding Al from the top to form a Pd/Al interface. Such is explained by cohesive energy differences between Al and Ru. 5) While complex mixed structures of oxides and metals cannot be predicted by simulation, proposals from STM data can be tested. In addition, model calculations expose basic factors such as interface preference, near-interface oxide film structure, and the resulting film stress, explaining the presense and variety of relief mechanisms. In this active field, examples, including experimental results, are drawn from alumina, magnesia, and titania. My valued collaborators are several. @footnote 1,2@ @FootnoteText@ @footnote 1@www.sandia.gov/surface_science/drj/. @footnote 2@Supported by US DOE.