Bert Frees and I have discussed this off-list to some extent. I think it is now time to submit some ideas to the wider community. Hyphenation is important in languages with many long words. One idea is to pre-hyphenate text before sending it to liblouis. This would be done by having liblouisutdml run a hyphenation algorithm, probably a version of the one now used in liblouis. This would generate an array similar to the present liblouis typeform parameter. This would be a char array and a bit would be set where a hyphen was permissible. This array would be passed to a new liblouis function called lou_translatePrehyphenated along with the usual parameters. However, it seems to me that prehyphenating every word is a lot of unnecessary processing. Hyphenation is needed only at the end of lines and perhaps in a few other situations. Since the inputPos array generated by liblouis is now accurate, it can be used to find the print word corresponding to the last word that does not fit on a line. This word could then be passed to lou_hyphenate for hyphenation and retranslation with hyphenation indicators as this function now does. However, the back-translation step in lou_hyphenate would be eliminated, since it is unreliable. After lou_hyphenate returns liblouisutdml would chose the most suitable hyphenation point. Thanks, John -- John J. Boyer; President, Chief Software Developer Abilitiessoft, Inc. http://www.abilitiessoft.com Madison, Wisconsin USA Developing software for people with disabilities For a description of the software, to download it and links to project pages go to http://www.abilitiessoft.com