bieling@terra.es wrote:
On May 24, 2:34 am, "Michael J. Mahon" <mjma...@aol.com> wrote:The introduction of the IIgs into the Apple II market is another example of this--"too little, too late".I think that the the Disk II is the reason why the Apple II's speed could not be increased steadily along the time, a la par with the contenders. A 'worse', less smart, but less 'synchronous' Disk II design may have permitted to speedup the the Apple II much more easily, as the CPU speed would not have been such an important factor in compatibility. Somewhere along the line, a timing source could have been added... if the need had ever arisen. Using the Apple II itself as an FDC instead of a Floppy Disk Controller *chip*, although clever and bright from an engineering point of view, has had some unforeseen, long-lasting (and bad) consequences...
But they were all rather easily dealt with by several different accelerator designs. The Zip Chip design was even licensed by Apple for incorporation in the //c+, so the path to faster machines was open by the mid-1980s. Another approach that would have worked well would have been to create a smartport 5.25" drive that worked for all non-copy- protected disks, with a "compatibility mode" that fell back to 1MHz-Woz-mode for old copy-protected stuff. Both approaches were straightforward in the mid-1980 timeframe. What made any simple "crystal swapping" approach to speeding the machine impractical was the synchronous integration of the processor and the video clock. That very clever design would have required a complete re-working to speed the machine up without a cache-based approach. The original Apple II design was a "crystalline" design, with everything synergistically dependent on everything else. This is both its greatest beauty and its inevitable "exoskeleton" that mandated a complete re-design to significantly improve it. If you combine that need with the need for very low-level compatibility, you have a recipe for getting pretty far from cost-performance optimality. Another fundamental problem was dependence on the 6502 line, which was extended in a very quirky way beyond 4MHz, without a second source and with spotty delivery guarantees. -michael NadaNet file server for Apple II computers! Home page: http://members.aol.com/MJMahon/ "The wastebasket is our most important design tool--and it's seriously underused."