I of course agree that "much work remains to be done"; but I am hopeful that instruction-execution counts will in time come to supplant CPU times, which are increasing problematic because no longer simply reproducible, for performance comparisons and evaluations.
It would be agreeable to be able at last to give literal meaning to the phrase "path length". If we must continue to use CPU times we shall all need to learn to think like agronomists, to give up point measurements and instead to view CPU times as mean values having high associated variances, and thus to recognize that explicit, formal statistical methods---experimental designs and multiple replications---will be needed to obtain meaningful results. John Gilmore, Ashland, MA 01721 - USA ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
