This must be something different from what we had at the beginning, as the constraints were for production processors commercially available in 1982-1983 and in either the Z-80 or 6502 architecture families. The reason for this is obvious. It has got to be possible to run the code in a real device in the "here and now" timeframe. Otherwise one could simply say well, "I wrote this code for the XYZ at 24.576 GHz and the whole program requires only one instruction." Further constraint should be that the author must own a running system capable of running the subject software on it in the native processor. This system must consist of components commercially available during or before the subject time period, i.e. 1983. Thirdly, it should be proven by replication (successful execution of the submitted code set on another computer not necessarily owned by the author of the submitted software) that this fete is achievable with the code set submitted. A simulator or even some fancier assemblers can be used to indicate the execution time. It's not desirable to include processors which don't exist yet in any comparison purported to be a timing comparison between Z-80 and 6502. That doesn't mean you can't look at whatever else someone might submit for general inspection and interest, but we do want to have a well-defined goal. -----Original Message----- From: Sellam Ismail <dastar(a)ncal.verio.com> To: Discussion re-collecting of classic computers <classiccmp(a)u.washington.edu> Date: Saturday, April 17, 1999 5:27 PM Subject: Re: z80 timing... 6502 timing dastar(a)siconic.com ---