"Peter Dimov" <pdimov@mmltd.net> wrote in message news:019c01c5256a$561efe40$6601a8c0@pdimov... | Peter Dimov wrote: | > But as I said, benchmarking hash functions is pointless. Real programs | > do not compute the hash value of a random 256 byte buffer five million | > times. | | ... or if they do, they use different 256 byte buffers and do not operate | from the L1 cache. ;-) no, but AFAICT we cannot exclude that a certain hash-function is much faster *and* otherwise behaves as good as the other "naive" ones. Can anybody answer this question: is it always optimal to use the whole range to compute the hash value; for eg. strings, if I know the average length of my strings, I can't see that it would make sense to process much more the the average length (or perhaps even shorter). maybe hash_range should be specified as temaplate < unsigned max, // max length of range to consider class Iter
hash_range( Iter, Iter ); ? -Thorsten