It is said that ISAM provide effiecient support for range searches, but i think it only works with the sorted files, because even the data are allocated sequential in ISAM, if the files are not sorted, how can it be effiecient?
with a sorted file: 4 8 13 index page
1 3 5 6 7 9 11 15 data page
in this sorted file, the range search would be effiencient since the data can be searched sequential from 1 3 5 6 7 9 11 15
but how about with a unsorted file? you can image that the search can not be effieciently since the data are not in order.