Started on the English 500 encoding correction. First wrote some XSLT to generate a correction-ready version of the file, by combining known-good data from the db-generated version of the file with supplied params and the student's incoming file. Then tweaked the poem.xsl transformation to make it more useful. Finally, worked on the first seven encodings, debugging the process as I went along, and sent those results on to AC. The rest should go much faster now.