From: William Morgan Date: 2005-04-23T13:02:41+09:00 Subject: Re: Question: Time efficiency of Array << Excerpts from Ara.T.Howard@noaa.gov's mail of 22 Apr 2005 (EDT): > yes that's true too. i've done tests in the past comparing bdb, > gperf, and other hashing type look ups against a bsearch and found > them to be quite a bit slower. on closer analysis it turned out that, > although hashing is O(1), the call stack is so much deeper for each > search as compared to a non-recursive bsearch that it predominates - i > was suprised. Just goes to show that theoretic bounds on worst case performance are sometimes just that.... -- William