Re: _Infinite Loop_ by Michael S. Malone

[Rubywand <rubywand@swbell.net>]

Supertimer writes ...
> 
> Rubywand <rubywand@swbell.net> wrote:
> 
> >> filled with the simplest componentry imaginable -- all superfluity having
> >> been winnowed out over the previous nine years
> >
> >     Pure nonsense. The IIgs is the most complex, least simple personal
> >computer ever built.
> 
> (See my reply to Dennis)
> 
> Are you kidding!

     Nope.


> I can plug in a card on the IIGS without worrying about
> IRQs or DMA channels.  I can plug in a drive without worrying about PIO
> modes or master/slave conflicts.  The IIGS is remarkably simple from the
> users' point of view.

     Yes; the GS has lots of nice features. On the other hand, it will not be
much good as an internet machine or for advanced graphics applications until
there is a major hardware upgrade.

     The PC, of course, remains bogged down by an old design which should have
been completely redone years ago. But, so far, no one company or group has been
willing and/or able to dictate a major platform revision.


> And the chip count was drastically reduced from that
> 9 years prior to the IIGS.

     Reducing chip count does not necessarily reduce hardware complexity. The
IIgs is a good example of fewer, bigger IC's and much greater complexity.



> Compare the chip count of the IIGS to that of
> the Apple II or II+.  The IIGS was elegant.

     In hardware terms, the IIgs is not at all elegant. It is a kluge, a
hardware design nightmare.



> The II and II+ generated heat
> almost as badly as a modern PC (I had a CPU chip failure on my PC
> recently, yuck).
>

     The II and II+ generate heat well within the capacity of the case/board
design to dissipate. Unless you add something like a SpeeDemon accelerator--
i.e. a plug-in poorly designed for Slot placement in terms of heat production--
the II and II+ need no forced air cooling.

 
 
> Maybe you mean from a design point of view.  It is true that the IIGS has
> lots and lots of circuitry in order to handly both the IIGS side and the
> IIe emulation side, but you have to admire the way the engineers (Woz
> was one of them) who designed the IIGS.  They took a complicated task
> and made it work so well, a complicated design on paper that is so simple
> in execution, the IIGS is simply the most beautiful piece of engineering
> Apple ever created.  They may deny it.  They tried to deep-six it.  But
> the IIGS was one of the finest Apple computers (that includes Macs) ever
> built.

     Forcing built-in, integrated old-II compatibility was a fundamentally bad
design decision. There was, simply, no need; and, the costs were enormous.

     Just as the first PC was a from-the-Intel-manual design with Apple
II-style Slots, a straight-forward native mode '816 design could have been
finished within 2-3 months of any 'Go' decision. There _could_ have been a IIgs
ready to battle the spiffy new Amiga's and Atari's in 1985.

     A pure native mode GS would have included new Slots supporting full
addressing. I/O would be out of the Zero Bank, "language card" kluges would be
gone, RAM expansion would be easy, and everything would run at the base speed.
For the same development and production cost, the GS could have started with
built-in stereo sound plus 320 x 200 256-color and 640 x 400 16-color video.

     Agreed, coming up with a hardware design which can also run most old-II
software was quite an achievement. Even so, it was a bad idea. As TrackStar
demonstrates, old-II capabilities would have been available within a couple
years as a plug-in option.

     We can learn from past mistakes. When the next major IIgs expansion is
planned, we will know to aim for simplicity and speed.



Rubywand