--------- START OF FORWARDED MESSAGE ---------

From System ....: Internet
From Conference : comp.sys.amiga.misc
MessageID ......: <335b3c9d.1092555621@news.jersey.net>
Originally from : Dave Haynie <dhaynie@jersey.net> <609176948>
Subject ........: Re: The new Amiga. Is it really coming soon ?

On 17 Apr 1997 19:27:26 GMT, gnoel@morgan.ucs.mun.ca (George Noel)
wrote:

>In article <5j3rgh$lot@proxy5.proxy.ozemail.net>,
>Colin Ward <colin@leprechaun.com.au> wrote:

>>  Yes, I know.  But the thing about the PIOS ONE is that it is too generic.
>>It doesn't have the custom chipset.  It's the custom chipset, *combined*
>>with AmigaDOS that made the Amiga what it is.  

In 1985, sure. But the custom chipset was an implementation detail. A
black box. It was the only way to get the Amiga's level of performance
back then at that price. You got something approaching the performance
of a 68020, graphics wise, back when 68020s cost about 1/3 the price
of the whole A1000. 

Today other folks custom chips have carried on this tradition. Just
ask any Cybervision or Picasso user. Again, it's a black box. All the
user sees is what the OS shows that user. How it gets done may be an
interesting technical detail, but the point is whether it's done or
not. Given that every still for sale today has significantly faster
graphics than AAA Amigas would have, much less anything that's out
now, this is not a point to be glossed over in our present state.

>You know, now I realize why a lot of Haynie's ideas at Commodore never
>seen the light of day.

That's less true that you think, especially you're talking "ideas".
All the early stuff I did eventually hit the market, if not from
Commodore, then later from others. After the A3000, I started doing
advanced projects. These were sometimes designed to be products,
sometimes just lead to projects. Such things as:
	
	BIGRAM The first Zorro III RAM board. 

	A3000+   This was a real project, but Sydnes killed it. The AA

		  work lived on in the A4000, A1200, and CD32. The DSP

		  work was brought back as a board-level product, 
		  which, though never released by C=, was licensed and

		  built by a third party for their own use. 
	Gemini	  The multiprocessor project, 1991. Killed by Sydnes. 
	Nyx	  The AAA system prototype, died with Commodore.
	Acutiator The A5000 system architecture. Some of this lives on
		  in the PIOS One. 
	Scarab	  Commodore's Zorro III graphics board. The Picasso IV

		  delivers the scan conversion architecture I designed
		  here. C= died before this could be built.

Anyway, that's the basic list from 1990 onward. Realize that the A2630
and Zorro III also began life in the skunkworks, they just managed to
be in the right place, and under a better management when production
time rolled around. 

I don't quite know the context of that comment, but it sounds intended
as a barb, though I don't know why. Don't barb things you don't
understand. There's some myth, I guess, that we sat on top of tens if
not hundreds of unreleased things. Not likely. Regular projects kept
most of C= engineering very busy; I got into more advanced stuff
because I really felt things needed pushing, and because I was working
with the chip guys as they came up with new chips, so the systems that
first got those chips were great places to try out new ideas. 

Dave Haynie      V.P. Hardware Engineering    PIOS Computer
haynie@pios.de     "...no RISC, no fun"

---------- END OF FORWARDED MESSAGE ----------