* Carlos <angus / quovadis.com.ar> [2004-12-19]:
> DICT = "/usr/share/dict/words"

....

>           stem = word.sub(/#{letter}/, "")

There seems to be a problem with german umlauts in words:

rq:16: premature end of regular expression: /?ge/ (RegexpError)
        from rq:15:in `each'
        from rq:15
        from rq:8:in `each'
        from rq:8
        from rq:7:in `open'
        from rq:7

Julius