How about creating a Open AMIGA Consortium?

[biggun]

Hello,

In the Amiga community there are still several driving forces
that are interested in developing and advancing the Amiga.

For example in the area of the Classic Amiga I see:
* The developers of the MiniMig
  Minimig is a great solution for A500 replacement.
* The Clone-A
  I think the Clone-A will be a A500 replacement as well.
* The Natami
  Natami as AGA successor. Focus of the NATAMI is AMIGA compatible as an A4000, but with more performance and enhancement like truecoler, 3D etc. And the Natami offer a CPU upgrade path.
* Elbox Dragon
  The Dragon is valuable CPU upgrade path (Coldfire).
  If the Dragon would be standalone with AGA included this would be a even nicer device.


I think that there is a lot of potential for these developers to cooperate.
The Natami team for example could lent AGA know-how to Elbox . Another very good idea would be to agree on a common enhancement standard.

I think it would be clever if the leading Amiga designers and developers would combine their brains and agree on a common chipset for the future Amiga.

For example: The Natami will include a "Pamela" (SuperPaula) Audio chip with more channels and more enhanced features.
The same audio enhancements could be loaded into the Minimig . This would allow to have a common Amiga nativa audio design. So applications using the new audio features could run good on all comming Amiga HW designs.


The key for success of the PC was that there was a open design. Any PC manufacturer could build a "compatible" PC.
I believe that it would be very clever to do the same now for the AMIGA.

Together we could create and specify a next generation classic AMIGA platform. I think its good that there are several people still interested in developing AMIGAs.
It would be a big advantage for the community if these designs help each other and guarantee that enhancements are compatibility to each other.

What is your opinion to this?

Does it make sense to create a AMIGA HW Consortium?
It think we should put the best brains together!

[biggun]

Yes and I tried to put together an "OpenAMiGA" spec, back in 2003-2004... with the idea of delivering a base specification of compatibility between the AmigaOS clones... no one was very interested and the idea fizzled out.

The idea is still good and the situation did change now.


In 2003 there was no new working classic HW available, was it?

This has changed.
There is the MiniMig, and Natami is working too.
We have AGA and even much more powerful now.


Today its not talking only but we can give parties something.
I believe its a big WIN for Elbox to join
as they could for example get AGA from it.


Cheers

[biggun]

I think we are mixing topics here.
My understanding of the topic was to promote the idea
of an open group of AMIGA HW engineers.

This open group would help to build genuine Amigas.

My definition of genuine,original, or classic Amigas
are machines that can run orignal Amiga games or apps, natively.

This definition of Amiga includes the following machines:

A1000 (OCS)
A500  (OCS)
A600  (OCS)
A2000 (OCS)
A3000 (OCS)
CDTV  (OCS)

A4000 (AGA)
A1200 (AGA)
CD32  (AGA)

MiniMig (OCS)
Natami (SuperAGA)


Some people raised the valid point about software APIs-
Sharing the software API with Neo-Amiga operating System
is a valid but complete different topic IMHO.
We should not mix this topics with the HW discussion.

At the time when MOS and OS4 where created no one was able
to create a system like the MiniMig or Natami.
At this time it looked like it will never be possible
to build new AMIGA HW and the best option available
was to use PC hardware and to run a AMIGA like OS on the PC.

The Pegasos and the AmigaONE are systems that are able to run a NeoAmiga operating system.  The Aone and Pegasos are on the Hardware level not Amigas but PCs.
They are no more AMIGAS as any Dell or MAC is.
There is nothing wrong with building a PC and runnning a Amigalike OS on it!

I think is a valid option to go this route, especially as it looked as new AMIGA he could never be build
Fact is that the situation has changed now.

Two people have proven us that building new AMIGA HW is possible.

MiniMig and Natami are true AMIGAs.

There are still companies interested in building classic AMiGA HW.
Elbox is working in this area ince years, their problem
is that the can not produce AMIGA chipset themselve so ELBOX expansion cards always require an orignal AMIGA (ie A1200)
If ELBOX would use the NATAMI chipset instead of the A1200 then they could creatre new, standalone systems a lot more powerfull than the A1200 based machines.
There are other companies which want to create new classic AMIGAS too.

I think we all agree that there is a lot potential of working together.
If you are able to create a AGA chipset the next logical step is removing some limitation.
- For example, you will want to speed it up.
- And you can add new features like truecolor or 3D acceleration -  As the Natami does.

Three companies can develop their own AGA enhancements or they could work together and agree on common things.

Three companies could work on integration the Coldfire into AOS. Or they could work together on this.

I'm probably working to long as open source developer but my honest opinion is that sharing and giving is more important than "taking".

[biggun]

Where's ECS in that list?

IMHO, ECS is a minor improvement that its not worth mentioning. But you can decide yourself.

In a nutshell the Chipsets have these features:
 

OCS
---
LOWRES
(320x256) 6-Bit (Planes)  = 64 Colors
HIRES
(640x256) 4-Bit (Planes)  = 16 Colors
Sound:
4 Channel x 14 Bit

ECS
---
LOWRES
(320x256) 6-Bit (Planes)  = 64 Colors
HIRES
(640x256) 4-Bit (Planes)  = 16 Colors
Productivity
(640x480) 2-Bit (Planes)  = 4 Colors
Sound:
4 Channel x 14 Bit

AGA
---
LOWRES
(320x256) 8-Bit (Planes)  = 256 Colors
HIRES
(640x256) 8-Bit (Planes)  = 256 Colors
Productivity
(640x480) 8-Bit (Planes)  = 256 Colors
SuperHires
(1280x256) 8-Bit (Planes)  = 256 Colors
Sound:
4 Channel x 14 Bit

Super-AGA
---
LOWRES
(320x256) 24-Bit (Truecolor)  = 16777216 Colors
HIRES
(640x256) 24-Bit (Truecolor)  = 16777216 Colors
Productivity
(640x480) 24-Bit (Truecolor)  = 16777216 Colors
SuperHires
(1280x256) 24-Bit (Truecolor) = 16777216 Colors
SuperHires-Productivity
(1280x1024) 24-Bit (Truecolor) = 16777216 Colors
Sound:
8 Channel x 24 Bit

Blitter: 100 times faster than AGA BLitter
Extras:  Enhanced 2-D Acceleration, 3-D Acceration



Yes, ECS was an improvement for WB users but not for games, I think that ECS is not that important.

AGA was of course a big improvement

And SUPER-AGA is a big improvement

[biggun]

bloodline wrote:
Actually, that's:

Audio: 4 x 8bit

or 2 x 14bit (trick).

It depends on how you look at it.
The quality of the analog output signal is always 4 x 14 bit.
So technically the OCS was supporting 4 channels with 14 bit quality.

But the 14 Bit has devided into a volume and sample value per channel.
As the Audio had 4 DMA channels you could not DMA load 4 x 14 bit.  

[biggun]

bloodline wrote:
Yes... that's why we need to focus on AROS, which is the only resonable future direction for the community, if the community wants to control it's own future...

I agree with you.
AROS seems to me as well as the most sensible way forward.

I respect the major efforts and excellent work that was put both into MOS and OS4.
But that MOS and OS4 are closed source is a very high risk.
Beeing closed source makes them very "fragile" to getting knocked or never updated anymore if the developer decides to  do it.


AROS seem to be much more robust against these type of risk.

[biggun]

The idea of the NATAMI is not to have the most fastest hardware on earth, but to make the HW  powerful enough to be able to fulfill the everyday tasks and to really make USE the HW.

Here is an example application running on NatAmi:

[biggun]

Likewise, the OS required to support this modern hardware scales up quickly in terms of complexity and needed resources. Look at linux.

I fully agree. Linux has of course many nice things.
But some of its "modern advantages" are paid with by serious complexity and performance issues.

If on Linux a application uses a device then there is for security reasons no direct connection between the application and the device.
This means that if a user application wants to write something to this device it will write this in its user space buffer. The kernel is than called to copy this user space buffer into the kernel space device buffer. The same complex handling is done when the packets return form the device.

This way is "more secure" of course but also twice as slow as the AMIGA solution.
We are writing 10G Ethernet drivers at work and the way Linux handles this is seriously impacting performance.

The beauty of the AMIGA solution is that its more elegant, simpler and faster!

[biggun]

persia wrote:
The retro-future version is more do-able.  [...] Allow use to use modern parts, USB keybords, mice and joysticks.  Improve the graphics so they are a little less retro.  

So you are proposing to use the NATAMI, are you ?
Have you seen: http://www.natami.net/

Its backward compatible to A1200 and A4000 but allows to use USB, Ethernet, oldl and new Amiga OS friendly truecolor resolutions with 1280x1024 and 24bit audio.