XR was faster than SR on some System/360 models, for the obvious reasons that others have already noted. The difference was never, however, very great or important.
Chris Blaicher's point is the important one now. It is sequences of instructions and the algorithms they embody that need to be evaluated. Comparisons of single-instruction timings can be, often are, radically misleading. Retreat into the old and familiar as a device for denying or avoiding complexity is, I suppose, understandable; but it is not productive. 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
