* This message forwarded from area 50, csa.advocacy
 * Original message dated Wed 17 Jun 92  2:46, from Skip@tacky.cs.olemiss.edu
(skip Sau

Newsgroups: comp.sys.amiga.advocacy
Path: news.UVic.CA!ubc-cs!destroyer!gumby!wupost!ukma!nntp.msstate.edu!tacky.cs
.olemiss.edu!skip
From: skip@tacky.cs.olemiss.edu (Skip Sauls)
Subject: Re: The Amiga is being left behind...
Message-ID: <1992Jun16.190007.27879@ra.msstate.edu>
Sender: news@ra.msstate.edu
Nntp-Posting-Host: tacky.cs.olemiss.edu
Organization: University of Mississippi, Dept. of Computer Science
References: <1992Jun8.204331.20944@ra.msstate.edu>
<1992Jun8.231850.9538@athena.mit.edu> <1992Jun9.084014.23701@news.iastate.edu>
<1992Jun16.111734.16163@mnemosyne.cs.du.edu>
Date: Tue, 16 Jun 1992 19:00:07 GMT
Lines: 140
In article <1992Jun16.111734.16163@mnemosyne.cs.du.edu> jyoung@isis.cs.du.edu
(Jeff Young) writes:
>In article <1992Jun9.084014.23701@news.iastate.edu> barrett@iastate.edu (Marc
N Barrett) writes:
>>In article <1992Jun8.231850.9538@athena.mit.edu> fihsu@athena.mit.edu (Hsu
I-wei) writes:

>>> The custom chips distinguish the Amiga, and the AA chipset would truly
>>> help convey the advantages of having these custom chips.  I hope there
>>> will be some midrange model (ie A2000+), something for A3000 owners (ie
>>> me) to upgrade to (A3000+), and a nice 040 Amiga(A4000) to top off the
>>> line for now.  Something like an A2000+ but with limited A500-type
>>> expandability would be a nice low-to-mid end thing.
>>
>>  I have serious doubts about these new machines (if there really is more
>>  than one) based on this mythical chipset.  Will Commodore bother trying
>>  to sell the new machines to people other than old Amiga owners?  Or will
>>  they market the new machines exactly the way they market the A3000?
>
>   Commodore said NEW AMIGAS will be sold soon so many developers have
>   A4000 and A3000+ in hiding?!  Can people say if they are keeping A4000?
>   If developers do not have A4000 NOW how can C= sell machines soon like
>   C= engineering manager said???
>
>   People with A4000 STAND UP AND SPEAK OUT!!

I've been chomping at the bit to speak out and right now I just don't care
about the non-disclosure agreements.  We have been beta-testing the A4000
and another model which may be called the A1000 Classic, A800, or possibly
some other name.  Both are based on the AA chipset and should be able to
hold their own against other machines on the market, but there are still
a few areas where they are lacking.

The A4000 is pretty much an A3000 with an 040 on the motherboard, the AA
chipset instead of the ECS, and beta 2.1 ROMS.  Ours has 16M fast RAM and
4M chip, expandable to 64M and 16M respectively.  ZIP RAM like that in the
A3000 is used, but there is a chance that production machines will go to
SIMMS.  Our machine has a 105M Quantum hard drive and a true full-speed
high-density floppy.  The expansion slots are Zorro III and appear to be
identical to those on the A3000.  I am not positive, but the cpu slot looks
to be the same as the one found on the A3000.

The other machine is housed in pizza-box style case 3" high and 15" wide
and deep.  It has a 16MHz 030, the AA chipset, 2.1 ROMS, and sockets for
up to 16M fast RAM and 4M chip RAM.  It has the same floppy drive as the
A4000, but the hard drive is only a 52M Quantum.  There is a daughterboard
with two Zorro III slots and one inline video slot, but it looks to be
very cramped in there.  The cpu slot is in the same orientation as that of
the A3000 and A4000, but there is not much clearance.  It is a very nice
compact package and manages to look even "sexier" than the A3000/A4000.

Both machines sport the same I/O ports, starting with the same ports found
on the A3000.  The floppy drive and serial ports are now stacked in the
same way as the SCSI and parallel ports, as are the RGBA, VGA, and audio
ports.  There is a microphone port that works with the built-in digitizer,
with 8bit 22kHz sampling possible.  Not exactly CD quality, but better than
the toy CODEC samplers found on the NeXT, Mac, and others.  The keyboard and
mouse ports are on the rear of the machine, with the joystick ports stacked
vertically.  This stacking leaves just enough room for a pair of MIDI ports
and another port that was covered with a plastic cap on both of our units.
We pried it off and it looks like a PCMIA slot, but we've not been able to
confirm this.  Given that the A600 has one, this would not be a bad idea,
but it seems odd that it was covered up like that.

The new mouse looks to be the same as the pregnant mouse, but the A3000 came
with a three button and the other machine a two button.  Both had standard
A3000 keyboards.  The most shocking thing about these machines is how much
they feel like existing Amigas instead of a new breed.

The AA chipset is a real improvement over the ECS, but it may not be enough
to keep the Amiga competitive for another 7 years.  Some of the chips are
similar to those in the ECS, but none look to be drop-in replacements for
existing machines.  Agnus, Denise, and Paula have been improved and all
three are now surface mount chips roughly the size of the fat Agnus.  There
are three other chips that have tape covering the names, but one of them
is certainly an Amber type de-interlacer chip.  It has what 1M of VRAM in
order to handle the increased color depths and de-interlaces all but the
highest resolution modes.  We peeled back the tape a bit on the other two
and one is named Lisa, the other was just  a number.  The Lisa chip appears
to be responsible for the graphics, but its relationship with Agnus and
Denise is not so clear.  The other chip may be the rumored DSP chip, but
none of the software we have mentions it specifically.  There is an A/D
chip for the microphone and two D/A chips for the audio in the same general
area, so they are probably related.

Amiga owners will be surprised at some of the features of the AA chipset
and disappointed with others.  The resolutions supported are similar to
those of the ECS, with the addition of the 800x300, 800x600 interlaced,
1280x480, and 1280x960 interlaced.  The Amber de-interlacer works with
all of the modes, but most SVGA monitors can't handle the non-interlaced
1280x960 mode.  All modes can be up to 8 bitplanes except for the 1280
modes, which are limited to six.  The palette is 24 bit, but there is
a way to simulate 12 bit for compatibility with some older software.  A
pleasant surprise is the SuperHAM mode which uses 10 bits to allow for
pseudo-24 bit color.  This works in the 320 and 640 modes and is nothing
short of spectacular, with the bonus of being able to handle up to 30fps
for some amazing animation.  The 800 and 1280 modes have 8 bit 256k color
and 6 bit 4096 color HAM modes, respectively.  There is still a problem
with fringing on raw images, but the better programs seem to do a pretty
good job with reducing the problem.  The sprites now change to match the
resolution of the screen, so pointers look much more professional.  They
may be up to 64 pixels wide, with no height limitation.  There are still
only eight per scanline, but the number of colors is up to 16 in all modes
except for 1280, which is only four.  All in all, these Amigas have graphics
that hold their own against offerings for the PC and Mac, but they won't be
able to last for another seven years without improvement.

We have not done much with the audio, but 16 bit four channel sound in
up to 56kHz is available.  A neat trick is that up to 8 8 bit channels
at 56kHz and up to 16 8 bit 28kHz channels can be emulated.  We don't
have any software that takes advantage of these modes with the exception
of the digitizer, but there is a lot of potential.

The 2.1 OS is pretty much the same as what has been posted on the nets,
but it really looks awesome on the new displays.  We have found the
800x600 256 color Workbench to be very nice and snappy enough to be
useable.  The blitter may have been improved, but it does not feel like
it is 4x faster as some like to claim.  There are only a few programs
that don't work on these machines, and all of those are pre-2.04.  If
the developers have followed the guidelines set by Commodore, they
should not have any problems.

There will probably be more than a few folks who are pissed off by me
letting the cat out of the bag, but it looks like something is needed
to keep more people from joining the mass exodus.  These machines are
stable enough for production, but the higher-ups seem to have their
heads up their asses and won't release them.  I guess I really should
not even bother trying to figure out Commodore's idiocy, but it is
frustrating to see such potential wasted.

In case anyone has not figured it out yet, this post reeks of being
a pipe-dream.  But then again it may all be for real.  It is left as
an exercise for the reader to determine the validity of this post.
Have fun.  :-)

Skip Sauls
skip@tacky.cs.olemiss.edu