How about creating a Open AMIGA Consortium?

[HenryCase]

persia wrote:
A custom chip is in you machine forever, if I want to upgrade to the latest Nvidia card it takes 2 minutes to open the case, pop out one card and pop in a new one.

Upgrading an FPGA is even easier, its all done through software.

persia wrote:
There's just no way a few hundred Amiga fans can outdo Nvidia

There isn't anyone seriously aiming to outdo Nvidia or ATI.

persia wrote:
Custom chips just create headache because some joker thinks they can write directly to them and then you can't upgrade because your software breaks.

Any bad changes made to the FPGA can be undone by resetting the Natami.

persia wrote:
There's really two roads here, a modern Amiga inspired by Classic Amigas or a reimplimented retro machine with a tweek here and there. There's a fundamental choice. You need to choice, abandon state of the art or abandon classic except as UAE.

There are more than two roads. Natami is in the middle of the two paths you proposed, being a classic/retro machine with substantial improvements. Probably easiest to think of it as an A5000.

persia wrote:
The sad thing is, no matter what the choice the community will grow smaller, and neither way offers a guarantee of success.

The level of success that Natami enjoys will be dependant on a number of factors, but it's certainly got a lot of interest from within the Amiga community and BBRV are showing an interest in helping out, so it may be sold at a reasonable price for ex-Amiga users to consider.

Sig999 wrote:
I think a consortium of sorts would be a good thing amongst the hardware developers. In a perfect world it could aid development by not having everyone reinvent the same thing... in a realistic world it opens the doors for them to share features - Super AGA takes off? No probs, you can plug this into your whatever other board..... everybody wins.

I agree, sharing the workload benefits everybody.

[A6000]

HenryCase wrote:
Upgrading an FPGA is even easier, its all done through software.
There isn't anyone seriously aiming to outdo Nvidia or ATI.
Any bad changes made to the FPGA can be undone by resetting the Natami.
I agree, sharing the workload benefits everybody.

If FPGA's can be reprogrammed 10,000 times or more then our new machines could be in a state of continuous evolution, becoming better every month or so.
Even users with no experience in VHDL could try their hand at making small improvements and if they work, submitting them to a library for others to try.
We may end up with something Nvidia would consider uneconomic to design let alone manufacture.
FPGA's mean our new machines will not become obsolete, but are FPGA's economically viable in large production runs, think big, be optimistic.

[A6000]

whilst it would be nice to allow users to tinker with their hardware as if writing a program in BASIC, we also need a manual write protection scheme to prevent hackers and viruses changing our machines for their own purposes.

[HenryCase]

A6000 wrote:
If FPGA's can be reprogrammed 10,000 times or more then our new machines could be in a state of continuous evolution, becoming better every month or so.

Certainly improvements can be made, but it would be better to limit these improvements to applications only, I don't want to keep on going through the process of upgrading my OS. FYI write/erase cycles should not be a problem. The Spartan-3AN, for example, offers 100,000 write/erase cycles. I have no idea what FPGA will be used for the Natami, but I doubt you'd see less than 10,000 write/erase cycles in its specs.

I gave an example earlier in this thread where you could have a custom FPGA core to speed up 3d rendering, and there are many other cases FPGAs could be customised for the benefit of applications.

A6000 wrote:
Even users with no experience in VHDL could try their hand at making small improvements and if they work, submitting them to a library for others to try.

That would be a great outcome of using FPGAs. Best start learning VHDL then A6000. :-D

A6000 wrote:
We may end up with something Nvidia would consider uneconomic to design let alone manufacture.

Absolutely, but understand that FPGAs are not going to be competing with Nvidia (or ATI) GPUs for pure power. As you say, there will be non-commercial applications that will benefit from FPGA-based hardware acceleration. Take a look at this video:
http://video.google.co.uk/videoplay?docid=8182954424526557863
As you can see a single FPGA makes the hacking app run 80x faster than on a 1.25GHz G4 CPU alone.

A6000 wrote:
FPGA's mean our new machines will not become obsolete, but are FPGA's economically viable in large production runs, think big, be optimistic.

FPGAs aren't as cheap as ASICs in large production runs, but the FPGAs used in Natami shouldn't cost too much.

[persia]

Maybe you are right, cutting edge is expensive and in the end you will never catch Apple or MS.  So abandon the cutting edge and go for the video game market with a machine that you can essentially reprogram the video and audio at will.  Price it competitively with Wii, PSP3 and X-Box.

But you have some serious problems, NTSC and Pal are of limited duration, ATSC will be the standard in the US next year with DVB replacing PAL faily quickly.  You need to write the OS to handle these.  Plus you need to read and play Blu-Ray disks and older DVD movies.  There's a lot of technology changes in the pipeline.  

[AeroMan]

Whateaver happens I'm in !!!!!
I believe we need to define directions of where to go from where we are, and I would like to be part of the design of an Amiga compatible.

So here goes some suggestions:

  - We could start some threads to brainstorm specific features of the machine, like the one about SuperPaula. Let's do it for video, CPU, Audio, I/O, etc...

  - We could compile the results in a set of documents and let them public. Keep some threads to improve then and freeze them after some time.

  - Start some development. Let's assign some tasks and have some wiring and programming :-)

  - Try some tests with Kick 3.1, and keep searching for AROS kickstart.

  - Leave space to use off the shelf components. Why can't we have custom chips and use commom parts at the same time? (I have some Ideas about it)

  - Glue everything togheter and create a new Amiga. Even if it can't be named Amiga due to legal issues...


Personal opinion:

   I don't think we should aim at a specific market like desktops or game consoles. If we define a nice spec, it could cover from a machine as small as a PDA or cell phone up to one as big as a high performance computer cluster.
   I seems to be just a matter of planning

[persia]

1) Kickstart roms in 2008 are nothing more than a dongle.  It serves no practical purpose in an open machine.

2) Designing scalable anything is hard, I seriously doubt anyone here has that talent.  So we are really going back to the cross roads, an Amiga that works in the corporate culture would require so many changes to the OS that it would break everything that currently works on the Amiga.  Even if we went that way do we have the necessary skillset within the community.  Remember this is a community that has yet to get OS4 to run on an old style Mac Mini, that's child's play compared to writing a fully functioning OS.

3) The reprogrammable FPGs are a nice idea but again the skill set necessary seems to be in the hands of 2 or 3 community members.  It would be far easier to use off the shelf components and limit the choices to say one companies video or audio card and then write drivers for them.  Maybe use the FPGs for a classic mode and a real video card for future development.  

4) Mobile phones require a lot of software to do the phone work, none of that exists in the Amiga world.

The skilled programmers are largely in the Linux world, how many hours went into making Linux into what it is today?  Realistically we are a small shrinking community that are mostly collectors or retro computing enthusiasts.

[AeroMan]

1)We don't need real ROMs. We need a software starting point, and KS3.1 is a good one. Another good choice is Linux stuff, but this way we would be just another distro... (better get a PC and install your favourit one)

2)Don't paint that as a monster... It is easier to provide a path from the scratch than reverse engineer two complex parts (Mac/OS4). To be scalable you have to specify your base and top model. For example, want a base ? This looks good:

http://www.freescale.com/webapp/sps/site/prod_summary.jsp?code=MPC512X&nodeId=0162468rH3DgbNGrmC22FA&tid=t32hb

Want a top ? Connect some of these through ethernet:

http://www.pasemi.com/processors/pwr1682M.html

Just think about what they have in common and how we could support then.


3)The base stuff needs to be in FPGA for compatibility with AGA. Off the shelf components are better supported in expansion boards. You can use common audio/video boards without having to design complex PCBs and have SuperAGA in FPGA locally.

4) The mobile phone was a small device example I used. It does not mean I would like these specs to support GSM. It means it should support simple devices. Take the PDA if it feels more comfortable, but you can choose anything small. The Freescale chip above would be nice in a PDA, but a G5/Radeon/Realtek combo not, so the base could be: "have a commom processor, drivers for audio and video", instead of "use a Radeon xxx and Realtek yyy". Sounds better ?

I know there are many people in Linux world and even more on MS. There are maybe less than 1% of that amount of people doing AROS and it looks really nice.

If everybody shoots in a different direction our future is as promising as the future of Sinclair users (no offense, I love my Sinclair  :-D).

It is worth a try. The Linux guys started like that. At least, we gonna have some fun !

 

[mike-]

@biggungunnar

I think its a great idea, one that really needs to be addressed, we could end up in a sitruation where we have 10 different systems all with 20 different "standards".

[amigadave]

bloodline wrote:
@amigadave

That's a very nice little fantasy... But it's not realistic, if the idea of an Amiga is to be real, we need to decide exactly what we want to acheive and why.

I've been here before 5 years ago, so I have no desire to rehash my old OpenAMiGA spec...

But think about it... A reasonable goal would be to create an opensource platform totally compatible with the Amiga500 with the intention of perhaps building an all in one games console... For example.

This is both achieveable and realistic, the minimig needs bug fixes and cost reduction, Aros needs the 68k port to be brought back uptodate (but we would only need 1.3 compatibility)... So there is a lot of work to do, but everything needed is already there. If this can be done then more advanced plans can be made.

My thoughts are a fantasy to most, but could be a possibility to many remaining Amiga fans.  Your idea could be a step along the way to something better.  My point is that we, the Amiga community, cannot and should not wait any longer for companies to come up with the next Amiga, or Amiga-Like computer hardware design.  With the advancements provided with projects like MorphOS, Minimig, NatAmi & CloneA, it is clear that it IS possible for the community to take charge of our future and move forward in spite of all the lies we have been fed by Hyperion and Amiga Inc.

Let's bring the community together again.  The Amiga was/is about all the so very talented, passionate, programmers and users that made/make the Amiga community the best and kept the Amiga alive long after C= folded.  It is time to unite under one common goal and work together.

[bloodline]

amigadave wrote:

bloodline wrote:
@amigadave

That's a very nice little fantasy... But it's not realistic, if the idea of an Amiga is to be real, we need to decide exactly what we want to acheive and why.

I've been here before 5 years ago, so I have no desire to rehash my old OpenAMiGA spec...

But think about it... A reasonable goal would be to create an opensource platform totally compatible with the Amiga500 with the intention of perhaps building an all in one games console... For example.

This is both achieveable and realistic, the minimig needs bug fixes and cost reduction, Aros needs the 68k port to be brought back uptodate (but we would only need 1.3 compatibility)... So there is a lot of work to do, but everything needed is already there. If this can be done then more advanced plans can be made.

My thoughts are a fantasy to most, but could be a possibility to many remaining Amiga fans.  

Well... anything is possible, but we do have to be realistic...

Your idea could be a step along the way to something better.  My point is that we, the Amiga community, cannot and should not wait any longer for companies to come up with the next Amiga, or Amiga-Like computer hardware design.  

I decided that 8 years ago when I put my, support and money behind teh AROS project.

With the advancements provided with projects like MorphOS, Minimig, NatAmi & CloneA, it is clear that it IS possible for the community to take charge of our future and move forward in spite of all the lies we have been fed by Hyperion and Amiga Inc.

I'm not sure what the MorphOS team are up to at the moment, but my point is that we have had the opertunity as a community to take control years ago. Opensource software and commodity hardware is the future. Opensource Hardware is good too, but you have to be realistic.

Let's bring the community together again.  The Amiga was/is about all the so very talented, passionate, programmers and users that made/make the Amiga community the best and kept the Amiga alive long after C= folded.  It is time to unite under one common goal and work together.

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...

[amigadave]

@bloodline,

Yes, I agree with most all you wrote, except I don't agree that AROS is the only answer.  

My problem with AROS is that the developers of it did not aim high enough and have still not integrated simple, no hassle, backward compatibility with the thousands of legacy Amiga applications and games. Those old applications and games are all we have now, other than a handful of programs and only a very few games that have been written for PPC only.

What I mean when I say did not aim high enough is, yes, I want some familiarity with the older Amiga API and interface, but there does not seem to be anything in AROS to excite and inspire more programmers to join and complete the project years ago.  Where is the "WOW" factor that makes anyone really want AROS.  I guess what I am trying to say (again) is that I want what the Amiga should have evolved into, if C= had not held back funding and further development that would have kept the Amiga ahead of the pack where it always should have been.

[freqmax]

@persia:
ATSC is a transmission and coding standard. NOT a video generator<->display link. Same goes for DVB.
Blu-ray etc.. can be handled via firewire/usb. No need to waste resources in a technological rat-race. I think you should read up on the technology issues.
Using of the shelf components instead of FPGA will limit the capabilities severly. And Amiga is not about mediocrity.

@AeroMan/persia:
I rather get a M68060+FPGA or PPC+FPGA (Virtex..). And then use technical skills to see how much performance one can squeeze out of it. The other way around is like.. we shall be no better than X and Y.

I think the idea to base an software API on the existing kickstart is a good idea to get a well defined API. That way we can easily find non-compliant implementations.

[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.

[persia]

You know I'm starting to change my mind on this whole thing.  Chasing state-of-the-art isn't really preactical, how long much effort to write code to take advantage of multiple cores?  How much to link machine to combine cpu power like xGrid?  Memory protection and all the other issues AmigaDos has are too hard to address.  

The Amiga is retro-computing at it's best.  I spend all day working on state of the art equipment.  When I come home and turn on my 2000 it takes me back to a simpler time.  Low end graphics, simple games and OS that really doesn't do anything but load programs.  It pulls me out of the 21st Century and back to my youth.  That's the Amiga magic.  Going state-of-the-art would ruin that.  The Amiga is like Amish furniture, nobody asks why the Amish don't use naugahyde!

[freqmax]

I no longer invest any resources in anything that can be taken away to will. At minimum the API should be available. And ofcourse even better the source code with a licence to change, redistribute etc.. (BSD/GPL).

Can't these be used as definition?
It will allow existing software to work. It's widely available, and almost anyone can verify compliency. Won't require any major changes as to how things work.

Hardware API definition: A500/A1200/A3000/A4000 etc..

Software API definition: KS 1.3 basic, KS 3.1 full.

Hw/Sw that works as above should be considered "Amiga compliant".