I can see that a little. It is a good article regardless what you drew from it. I found it more of a lossy set membership test that used packed cache lines to help in performance (over chaining), but I don't really know what his use case was. Good write up none the less.