Contents | < Browse | Browse >
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%% Musings and Analysis on the Amiga                  By: Aric R. Caley %%
%% Operating systems and a plan for the future        dances@qedbbs.com %%
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

        AmigaOS:  Is it worth it?
        -------------------------

  Many  people  say that AmigaOS has fallen way behind, and that it would
require  to  much time and work to get it up to speed with other "modern"
OS's.  I think they're wrong and here's why.

  Commodore  computers have historicly done well as home computers.  They
are  inexpensive  and  powerful.   Therefore, the OS can lack some of the
high  end  features  --  in exchange for being a small, efficient OS that
runs well on a home computer.

  Other competing OS's have their share of faults as well.


        The "competition"
        -----------------


Well lets look at Ms-Windows, shall we?

  It  still does not have pre-emptive multitasking, though version 4.0 is
supposed to have it (but it's not out yet).

  It does not have memory protection.

  It  still  uses  that  rediculous  MS-DOS format for their drives, even
though  it lets you use names greater than 8 characters it must cram them
into the old format.  Quite a hack.

  It is slow and memory hungry.

How about the Mac's system 7.whatever?

  It  also  still  does  not have pre-emptive multitasking, and who knows
when it will get it.  A massive rewrite will be required to add it.

  It also does not have MP.

  It's also slow and memory hungry.

Of  course I'm being very brief here, but I'm just illustrating that they
lack some important features too.

How about AmigaOS itself?

  We have pre-emptive multitasking.

  We  also  don't  have  MP.  While it would be nice to have, it would be
very  hard  to  add.   Probably  as hard as it will be to add pre-emptive
multitasking to System 7.

  We're very small and very fast.

  We don't have virtual memory standard, but can add it.

  We don't have RTG or DIG standard, but can add it.

OK so we have some problems.  But I don't think they are insurmountable.


        AmigaOS's market
        ----------------


Consider the market for the Amiga.  It consists of:

  Home computer users.

  Game players.

  Specialty   applications:    Kiosks,   "set  top"  devices,  and  other
applications requireing an inexpensive multimedia delivery platform.

  Developers for all of the above.

  Video and multimedia.

  AmigaOS  as  it is right now is plenty sufficient for most of the above
applications.   It's  perfect in fact, because it is small and efficient.
I  hear Scala is trying to port their wares to the PC, and in order to do
so they are writting an AmigaOS-like custom OS for the PC.  Gee, what was
that  about  their  OS's  being better than ours?  Not for everything, it
seems.

  For  the  rest,  there  are many things that can and should be added to
AmigaOS.  So here goes.


        What to do with AmigaOS
        -----------------------

Several virtual memory programs have already been developed.  Simply pick
the best one, and integrate it into AmigaOS.  VM problem solved.

Look  into the many RTG systems in use by graphics cards.  If these small
companies can do it, the new Commodore should have no problem doing it as
a  standard part of AmigaOS.  Possibly parts of a current system could be
adopted.

Make  the  "post.library"  a  standard system library.  Make a Postscript
font  engine.  Make a Postscript printer driver than can print Postscript
to ANY printer (pass through for those that know PS).

Get  all the datatypes drivers that have been written.  Clean them up and
include  them.   Write  an  MPEG  datatype that can use the MPEG hardware
(CD32,  Zorro  card)  or  software.  How about some datatypes for all the
music   formats?   Some  more  text  Datatypes  would  be  awesome  (RTF,
Wordperfect,  etc).   And  a  Postscript Datatype.  And get developers to
support them.

Implement  something  like,  or get the rights to, Hotlinks.  How about a
Hotlinks  Datatype?  Would this not give us the equivalent to DDE and OLE
in Windows?

Create and publish more standards.  We could use some more standarization
of ARexx commands.  For instance, there could be a standard basic command
set  for  a  wordprocessor,  or  a terminal program, etc.  We need a good
standard for animation with sound/music.  People need to support DR2D (or
we  need  to create a better standard if necessary), and a formatted text
standard, through the clipboard.

We  need  more  robustness  in  the  GUI.   We  need  a standard for font
sensitive,  inteligent  GUI's.  Something both FAST and small and easy to
use  (which  effectively  eliminates  every  third  party solution I have
seen).

How  about getting in on the ANSI standards commitee that is hamering out
a  standard  for REXX?  After all, REXX on the Amiga is probably the most
popular  (or at least 2'nd behind OS/2) version in wide use.  Get William
Hawes   to   do   a   new  release  of  ARexx.   Put  in  those  proposed
object-oriented  extensions  I  read  about  in  the  REXX magazine (yup,
there's a REXX mag, it seems to only recognize the OS/2 REXX though).

Rewrite workbench, please.

Envoy should be part of the OS as standard.

I'm  sure  you  can think of a dozen things to add as well.  With all the
talent  out there, I think we could add a ton of features to AmigaOS very
quickly to get us up to speed.

-----

  The question is, is it too late for the Amiga?  Is there any life left?
Can  the  Amiga  be  saved  even by a company that knows what it's doing?
Ever  the  optomist,  I  think  it  can  be  saved.  I still think AAA is
something special, even though it has been delayed significantly.

So, how to do it.  I'm sure everyone has their own ideas about "the plan"
to save the Amiga.  I do too.  And here it is.


Step one:

  Restart  production  of  the A1200, A4000, A4000T and CD32 immediately.
Deliver  them  in  abundance  to  everyone.   I  hear there is a Canadian
company  that wouldn't mind buying 200,000 CD32's if Commodore could make
them.   Think of what that will do for the price of the CD32?  (for those
who  don't  understand  mass  marketing, it would reduce the price of the
CD32).

  Ensure dealers, suppliers and end-users that the Amiga is not dead, and
that  they are being manufactured to meet demand.  Let everyone know that
this isn't the old Commodore anymore, and that the new Amiga is not going
to just play around anymore.

  Help  developers make the decision whether or not to leave the Amiga --
give them every incentive to STAY.  Actively "court" new developers.

Slash  prices.

Ship all machines with 3.1, of course.

  Amiga model                   Competitor-target

  CD32  - $250-$300             Sega CD (haha), Jaguar, CD-I (HA!), 3DO.

  A1200 - $250-$300             Low end clones?

  A4000/030     - $800          
  A4000/LC040   - $1000          - Clones, Mac Quadras.
  A4000/040     - $1200         /
  A4000T/040    - $1500         Quadra tower?

Restart development of AAA and next generation Amiga, DSP card, etc.

Begin advertising CD32 in game mags and during children's prime time TV.

Advertise Amiga's as multimedia computers.


Step two:

(time frame: late 1994)

  Push  the price of the CD32 as low as possible.  Get CD32 displays into
stores,  to show off the games and the MPEG video (fight the CD-I head to
head  I  say!) Be sure to supply good quantities for X-Mas!  CD32 may not
be as powerfull as the 3DO, but it's half the price and has more games.

Get those A1200 and A4000 CD32 upgrade cards out!

  Get  software developers to support 3.1 features (because of course, by
now,  tons  of  people  should  have the cheap 3.1 upgrade kits, right?).
Write  an  MPEG  datatype that uses the CD32 MPEG card if you have it.  I
want  to see multimedia software running in a window on my workbench with
real  smooth video in it!  I want to be able to cut and paste any kind of
data from any application to any other!

  Bundle!   Pick  a  good  gfx  card and bundle it with "hi-end" A4000's.
Pick  a  good  040  accelerator  (and  060 when available) and bundle it.
Bundle  GigaMem.  Bundle Ethernet cards and Envoy.  There's lots of third
party  stuff  that  is  good,  why  not support them (just stay away from
GVP!).

  The  bundling  could be handled by distributors; cooperate with them by
unbundling Amiga's (sell them motherboards and other parts, separate from
each  other  so  they can put together custom systems cheaply).  I'm sure
CEI and Creative Computers as well as individual dealers would love this.

  Amiga model                   Competitor-target

  A4500/040-40  - $2000         Quadra 650, 800.
    (40Mhz 040, gfx card)
  A4500T/040-40 - $2300         Quadra 950
  A4500/060     - $2500         PowerMac 7100
    (66Mhz 060)

More  advertising:   multimedia!  Imagine for example the following print
ad:

  "Multimedia does not live by CD-ROM and cute speakers alone.."

  Many people would have you believe that you can do Multimedia by simply
  adding  a  CD-ROM and a sound card.  Well, Multimedia is much more than
  that.   Multimedia means fast animation, great sound, video, hypertext,
  and  interactivity.  In order to do all these things, you need a system
  designed from the ground up for Multimedia.

  The Amiga is simply the best Multimedia platform.  From the inexpensive
  CD32  unit (perfect for Kiosks and consumer use), which costs thousands
  less  than  other solutions, to the Amiga 4000, the Amiga can't be beat
  for  the  price.   With  the  Amiga line, you can reach anybody in your
  target market.

  Not  only is it a great delivery platform, it's also a great Multimedia
  development  system.   With  advanced,  award winning products like the
  Video Toaster (3D animation, video processing, and more) you can create
  spectacular Multimedia presentations.

--

You get the idea.

Hopefully, X-Mas '94 will be profitable, with very little investment.

Step three:

(time frame:  early '95)

Work closely with Amiga developers to produce products for the Amiga.

Crank  out  as many CD32's, A1200's and A4000's as will sell (worldwide).
Get CD32 into toy and department stores, and A1200+CD32 expansion bundles
into the higher-profile places.

Release the DSP card (Zorro III).

Release  a  revised A4000:  better memory system, supports more than 16MB
RAM,  SCSI-II,  DSP  in  "DSP  slot".   Compact case, full size and tower
versions,  replace  all  previous  A4000's.   "The A4000 that should have
been".  "Multimedia workstation".

Release revised A1200:  fast RAM SIMM socket, math co socket, clock built
in,  Akiko  (CD32  chip),  28Mhz EC030.  $350-$400.  Add CD-ROM for $150.
Might  as  well  do  a  CD32  with the same processor..  "multimedia on a
shoestring", "Advanced gaming pleasure!".

  Amiga model                   Competitor-target

  CD32          - $200
  CD32-II       - $300          3DO, Jaguar-CD

  A1400         - $400          Low end clones?
  A1400CD       - $550          3DO?  Jaguar-CD?  :)

  A4200C        - $1000         Quadra 605/650
    (33Mhz 040. DSP optional)
  A4200/040-33  - $1200         Quadra 660AV
    (DSP)
  A4200T/040-40 - $1500         Quadra 840AV
    (DSP)

Step four:

(time frame: mid 95)

Aggressively   market   the  Amiga:   Multimedia  and  Games.   A4000  is
multimedia,  CD32 is games, A1200 is the cross-over machine.  Keep prices
low!   Get  people  to  use  Amigas.   Get people to sell them.  Get more
developers.

Aggressively develop AAA and new Amiga:  Pheonix 5000.

Aggressively develop AmigaOS 4.0 (see my other article about AmigaOS)

Make some profit?

Step five:

(time frame: late 95, BUT BEFORE XMAS!)

Rollout  Amiga  Phoenix  5000 series.  Work with NewTek to simultaneously
release an AAA based Video Toaster/Flyer(!!).  "Toaster 5000"?

* AAA chipset (see my previous article about AAA)

  Briefly:

  1280  x  1024  max  viewable  display (non interlaced; at least 8 bits,
    maybe 16 max).
  Very fast blitter, can do mathematical operations (image processing?)
  32/64 bit DRAM or VRAM depending on configuration.
  upto 16MB chip ram.
  Planar (1-16) or chunky (2,4,8) pixels, 24 bit modes, compressed mode.
  8 channels of 16 bit audio.

  Now  I've  heard  people  whine  that  "1280  x  1024 sucks, it MUST do
better!".   Well, have you checked out the prices of 21" or larger colour
monitors?   They  ain't  cheap.   If you can afford one of those, you can
afford  to  get a fancy (expensive) graphics card that can do better than
1280 x 1024 if you really need it.  The standard graphics aren't going to
be able to cover ALL cases of what people need.
  The  Amigas'  niche  is  video,  for better or worse, and AAA is plenty
sufficient  for  video  in  the conceivable future (it shouldn't have any
problem even with HDTV resolutions -- something that other graphics cards
WILL have a problem with).

* DSP (possible multiple units)

* 40Mhz 040, 50/66Mhz 060 (with 64 bit localbus)

* SCSI-II

*  PCI  only  on  "low  end"  systems, with additional Zorro III slots on
mid-high end (ZIII slots would be faster than A4000's).

*  No CPU slot -- CPU uses PCI slot.  In theory, you could have more than
one CPU card.

*  Aggregate  system bandwidth of 400-600MB/sec (compared to around 35-48
for an AGA machine).

*  AmigaOS  3.5  (will  probably  be  preliminary,  a  bit buggy, and not
available  for  older  Amiga's  till  4.0  in  '96).  Software to use DSP
(VCOS).  Networking, Multimedia extensions (Envoy, CD32 software).

  Amiga model                   Competitor-target

  A5000-060     - $?            PowerMac 8100 (or whatever comes next)
    (8MB fast, 2MB chip)

Continue  with aggressive multimedia marketing, with the same theme (that
CD-ROM  isn't  all you need for multimedia!).  The P5000 should be pushed
as simply the ultimate multimedia computer.

--

OK,  so  maybe  this is overly optimistic.  It may take a bit longer than
what  I outlined above.  But I still think the Amiga has a chance.  We're
just farther behind than we should be.  It still might be a bright future
ahead.

-----

[To preserve Mr. Caley's opinions and feelings, I have left this article
largely untouched.  However, it is worth reminding that no purchase of
Commodore has been completed and that the architecture possibilities of
future machines vary, depending on who you talk to.  AAA may not be
a viable future option-instead, another chipset may take its place. -Ed]

Bio:

Aric  Caley  is  Amiga  programer  of 7 years.  He is currently attending
Cypress  Community  College in California working towards a BA in English
(with   a   minor   in   Computer   Science).   You  may  EMail  him  at:
dances@qedbbs.com.