Closed samebchase closed 11 years ago
@samebchase I tested with
(time (evenp (qtl:range-product 1 200000)))
on SBCL with no optimizations and full optimizations, with and without your patch, and I only got about 0.005 seconds difference.
Private discussions lead to the conclusion that the change did not speed things up. Closing.
old: mid = (upper + lower)/2 new: mid = (upper - lower)/2 + lower
Rough benchmarks with time indicate this is twice as fast