Needed a way to identify words used as examples in the annotations, so added the "mentioned" element. Entailed - adding the element to a data file - invoking oddbyexample to create an odd file incorporating the new element (see previous post on that) - using Roma to generate a new rng file from the odd file - putting that rng file back into my tree - updating the xslt files to find the new element correctly (special case was mentioned embedded in note) and then output appropriate html - committing the changes to the repository and uploading the modified files into the exist db. For the time being, we're using it for any string that we want italicized as the vast majority (if not all) are in fact mentions. If we find a significant number of other cases of strings that need italics, we'll make the necessary modifications.