When adding a third sample text to examples/KP.pl, I saw three consecutive lines with hyphenated (split) words, including the penultimate line. It is my understanding that the K-P algorithm is supposed to avoid such runs of hyphenation (as well as hyphenating the next-to-last line), so I'm going to consider this a bug.