The question: "We are relatively new O-users, and we are using it to rebuild our structure and add waters with higher resolution data. We are also fixing up bad side-chain rotamers. My question is, when we move parts of the structure around with move_atom or tor_residue (and save the new coordinates with "yes" "yes") the atoms seem to revert back to their old positions. This is even after we've used save_db or sam_atom _out. What could we be doing wrong?"

If you say "yes" "yes" you're probably using move_atom and tor_residue simultaneously; I don't think this works (by the way: move_atom is the worst possible command one can use for rebuilding amino acid residues!! Use lego_side, move_zone, tor_res; mo_frag in rare cases; rsr_rigid if you have high-res data).