On Thu, 20 Aug 2015 18:46:30 -0700, Charles Mills wrote: >No, you only have to save the hash. > *Provided* that you abandon Kirk's idea of optimizing by a quick preliminary hash of a small fraction of the data and hashing the entire data set only when the subsets are equal.
I suppose you could save both the short hash and the full hash. Does the benefit justify the complexity? -- gil ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
