Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!ncar!boulder!tramp!hassell
From: hassell@tramp.Colorado.EDU (Christopher Hassell)
Newsgroups: comp.sys.apple
Subject: Re: Future of Apple II, wishes [Listening, Keith???!!??]
Message-ID: <6393@boulder.Colorado.EDU>
Date: 5 Feb 89 07:03:07 GMT
References: <458@madnix.UUCP> <1164@mentor.cc.purdue.edu> <25312@apple.Apple.COM>
Sender: news@boulder.Colorado.EDU
Reply-To: hassell@tramp.Colorado.EDU (Christopher Hassell)
Distribution: usa
Organization: (Let's see, I'm positive <clang> .. I've <rustle> got ... )
Lines: 67

I've had a few requests for my old wild suggestion as per GS+ AND since
everyone here is musing about a New GS or Whatever // you like....

My vote goes beyond that of simply giving a coprocessor..

I say schlunk a WHOLE NEW PROCESSOR down there <and that's it!!>.  

The idea is basically this:  <Keith, being that you have become a bit vocal
                               lately [ thank you, too ] Please Followup and
                               say what you think of this>

   - One can be the "compatibility box", number cruncher, general
        quick, dependable, and interactivity-handling speed-demon.

   - The other (specifically NON-equivalent but STILL programmable if not 
                                                         command-set compatible)
       would be a constantly-interrupted, prefetching, buffering, background-
       -handling, generally-doing-stuff-that-special-chips-might, and also a
       NEARLY COMPLETELY DIFFERENT programming environment.  Possibly even
       supporting true multi-tasking???  (Talking REAL expandibility possible)

    The hardware could handle:  
         Some dual-ported RAM for communicating  [possibly multiplexed too]
            <possibly even write-only and read-only too>
         General "locking" of slots to writing for one processor 
            <Note writing, ONLY.  Reading could still be by the other,
               a mask could also be put in place w/regard to soft switches>
         Using the already-existant horiz-line interrupts for updates.

         etc... many other means of harnessing the power of stuff, while still
           being compatible.

My reasoning is basically:

    Generally speaking, a survey showed that only one background-process is
      needed for personal computing, more was considered a luxury.
 
    Context-switching is all that normal types of parallelism (or multitasking)
      really need, so that need only be determined by size of memory or an MMU.

    MANY tasks are NOT necessarily required-finished before another thing can
      start.  This just BEGS for parallelism (read graphics BTW!!!!!)
    
    LOTS of things could be performed by a mildly fast processor that would 
      take ACRES of special-chips otherwise.  <Software-sprites!!  Music stuff>

    Parallelism could produce SERIOUS software-abilities that were made for it.

    Given upgrades, people wouldn't have to depend on the Megahertz fairy to
      fix up their computer constantly.
  
    Connectivity is The Next Wave as things are AND ANY SORT OF 2nd processor
      would just LOVE handling the overhead for the network, and or modem,
      and or file-server ..... ad exaltum.

Two has the AMAZING quality as a number of being one MORE than one.
This allows for so many things to be possible I could just scream.
I DON'T want to hear tons of people talk about how tough it is to use
  more than one processor at a time, boo hoo.  DANG if you does it right
  there's NO reason they need interfere with each other at all.
--------
This may be an upgrade or MORE down the road but DANG if it wouldn't be THE
MOST GUTSY AND USEFUL APPLE CORP. AND COMPUTER I'D EVER SEE!

But.. don't take my word for it.. Let's hear an Apple Person hisself talk
about how impossible this is.  (I suppose it might...but I know everything.)
### C>H> ###