At 10:13 AM -0400 4/28/02, Hans Geuns-Meyer wrote: > > I replaced my hashing function with the universal hashing function >> and the program works fine, but I am still finding that there are a lot of > > collisions, and I am not sure how to change that.
Another lesson to take away is that linear probing is not the best. The other two schemes described are quite easy to add. Given the amount of instrumentation you have, it would be easy to check to see if one of the other schemes performs better. - jeff parker