what if they get out of order?
i had to do a fuzzy search on medicare/medicaid files for a state before, after dumping a few gigs of cobol tape onto a disk. doing a crunch on the data searching for duplicates, on a fuzzy search mind you (partial name matches, date of birth, miskeyed id's, and about 2 other fields) that took about 2 hours to run on the full db. bout 1.2 million entries, had to run 6 macros and create linking tables, all that crap. i like data mining.
