In particular, it was very nice to notify the user which,
specifically, keywords were not searchable. Chris' wall-clock times
were also amongst the lowest, and his looked to be the most
portable...i think that i could apply it to a document database with
any number of concepts, different sizes of loads, and with some
concepts not appearing in the database at all. Very nice.
and his documentation was exceptional; i had a crystal-clear
picture of how, exactly, he implemented the FAST-INV algorithm. His
source code was very short and efficient. (jealousy)