Excerpts from Michał Marczyk's message of Tue Feb 23 14:54:10 -0500 2010: > That's a start. Then you'd also have to take care of key collisions.
Yup. As far I can tell, this would be easy, and hopefully not too much performance problem in the no-collision case with unboxing. > Incidentally, there's no point in hashing word-sized integers. > (They're normally used as unchanged as their own hash values for > purposes of insertion into hash maps.) If I recall correctly, hashing integers tends to be useless except for certain pathological distributions which cause all of the items to be put in the same bucket. I wasn't sure if Java had any protections against this; inspecting the source code, it looks like they just take the integer modulo the number of buckets. > I'm a bit confused now... Your initial numbers suggest there's not > much difference. Anyway, if your Haskell version doesn't perform very > well, perhaps it would be useful to ask some Haskell ninjas (from > Haskell Café maybe) have a look at it. I'm not one myself, > regrettably. :-) My initial numbers were shown to be wrong. :-) I haven't setup new numbers yet. I haven't gotten any bites from haskell-cafe yet. Cheers, Edward -- You received this message because you are subscribed to the Google Groups "Clojure" group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
