Hi, Ladislav,
if you are interested only in an associative array,
then a properly implemented hash should beat the
hell out of even a binary search on large data.
I guess it depends somewhat on the quality of the
hash function and other aspects of the implementation.
So, if you can beat the Rebol hash, which is probably compiled
c-code, with your binary search written in interpreted Rebol,
then RT had better take a look at their Rebol hashes.
Would you like to share some test source code with us
so we can try to duplicate and analyze your results?
Thanks!
-galt
p.s. Does memo-ize only make sense when there are a limited
number of possibilities for the parameters? If the space of possible
values is very large and requests could come in a sort of
random or rarely repeating way then the memo-ize trick wouldn't
do much good, would it?
