Re: _Infinite Loop_ by Michael S. Malone

[ChemSleuth@earthlink.net (Dennis Doms)]

In article <37470386.20F5FBFA@swbell.net>, Rubywand
<rubywand@swbell.net> wrote:

>     Pure nonsense. The IIgs is the most complex, least simple personal
>computer ever built.

I suspect the inference  was to the motherboard chip count (essential
features having been identified and integrated onto custom VLSI) ,
which for the IIe and IIgs was *very* low versus competitors at the
time. That integration resulted in fewer potential points of failure
and better reliability (except for a few design glitches like
inadequate cooling if too many expansion cards were added)..

>    "Stereo" sound output _should_ have been included; but, it was just
>enabled-- users have to build or buy a 'stereo board'.

Agreed. More heinous was the use of an inadequate internal speaker
(easily overdriven into distortion given the quality of the sound
circuitry that drove it)), which would have only cost a few cents to
fix (they used a better one in the Macs, of course).

>     A good joke! There was nothing "reasonable" about the price of the IIgs.

At introduction it *was* reasonable for a top-tier system. The problem
came when Apple never significantly adjusted the price over the years
(compare the price progression of the Mac SE to Mac Classic over the
marketing life of the IIgs), and (IMHO) obscenely gouged on some items
versus counterparts being integrated into the Mac (compare the cost of
a complete Mac Classic with internal hard disk and SuperDrive with the
cost for an external Apple hard drive or 3.5 drive required for an
Apple II of the same vintage).

It was introduced beside the Mac II, which cost several thousand
dollars more for similar capabilities (sound, color graphics, GUI OS,
etc.).

>     Hardware is a different ball game. The II/II+ looks like a machine
>produced by a 'hardware person'. There are a few significant flaws-- assignment
>of I/O addresses, fixed location of graphics pages in the middle of program
>space, and failure to build in provisions for simple 64k bank switching. Still,
>overall, the design is fairly straight forward and not all that difficult to
>expand upon.

Again, in context of the times (remember the design originated about
the time of the first S-100 machines, Commodore Pet, or original
TRS-80!) it was an amazing package. The hardware design was driven by
the high cost of components versus the low cost of programming around
them. That economic has since reversed. (At the time of the original
design 64KB of memory for a PC was a remarkable concept; at least the
Apple II supported using the entire address space of the processor!)

There is quite a bit in the first half of Malone's book about the
design decisions on the II, but a better sense can be developed by
studying some of the other designs of the time. (Also remember the
Apple II did color as standard, unheard of in a machine of that price
at the time, due to the unusual video design).

>     The hard-wired limitations of IIgs design more or less force going to a
>new motherboard in order to optain needed upgrades in memory and display. Since
>a new motherboard seems unlikely, we are left with a 'separate box' add-on
>approach in order to move onward to the next series of significant upgrades.

That also harkens back to the memory-mapped I/O model of the original
Apple II. And I'm also not sure it's a fair criticism since
motherboard replacements are forced on PCs due to the video
progression to higher-speed buses (from ISA to PCI to AGP) and the
various trends in memory (discrete chips to SIMMs to DIMMs). The IIgs
is no more limited for its time than the IBM PS/2 Microchannel
machines of the same vintage. (I have the headache of "transitioning"
a couple of hundred of those at work. :)