AAA Chipset in software?

[Gorf]

Originally Posted by Marchie

I can't deny the AAA board looks fascinating. Completed but never produced...

since the AAA chips were still full of bugs...

AAA was still supposed to be backward compatible, was it not? So exisitng software should work with it no?

fully compatible to ACS/ECS but not fully to AGA, since AGA development started actually later than AAA, as Commodore realized they couldn't finish it in time and resorted to something much simpler and closer to ECS.

that's why AGA has a completely unchanged Paula a almost unchanged Alice/Agnus and only Denise/Lisa got a real upgrade.

An FPGA implementation sounds cool, I know I'd try it out of there was ever a Vampire core with AAA. But hey, there's a big gap between 'I think it sounds cool' and 'someone should spend their time doing it'

Vampire is doing some more reasonable/useful updates like 16bit support in Pamela/Paula or SVGA and probably some 3D acceleration in future.

AAA would only be nice for historical reasons...

I would actually suggest to put in some features outlined in the Hombre specification, that are - at least in the video section - closer to the Amiga Chipset than I originally thought (documents where released some while ago)

I don't think a FPGA implementation is worth it, but in software it might be a interesting thing to play with.

[AmigaHope]

There's absolutely no way to accomplish this. Even if you had one of the Nyx boards, they don't even boot. We have no idea how Linda, Monica, and Andrea worked -- because they never worked.

It's not like the C65 where there are actual working prototypes that we can base our emulation on.

[quantum8]

AAA is an interesting anecdote in history and not much else since it wasn't released and nothing was written for it.

[Turrican_3]

Originally Posted by AmigaHope

There's absolutely no way to accomplish this. Even if you had one of the Nyx boards, they don't even boot. We have no idea how Linda, Monica, and Andrea worked -- because they never worked.

As far as I understand (see also Gorf's post) they were more or less functionally complete, but needed more time for proper debugging.

They definitely could be able to do something meaningful considering the Nyx/AAA prototype inside C= laboratories appears for a few seconds, turned on and with a familiar AmigaDOS screen on the infamous Dave Haynie documentary "The deathbed vigil"

Now obviously the question "is it worth it / what would be worth it for" is a completely different matter...

[Phantasm]

People put a lot of effort into things that aren't worth it all the time. I for example have put almost 2 years into reverse engineering and redeveloping the Ami-Express BBS software. It's something I have nostalgia and a passion for so I enjoy working on it even though there are only 2 or 3 people actually using it. You never know someone might just decide it's cool enough to develop even if its not worth it.

[grelbfarlk]

Ok how about this, lets flip the title to RTG in Software. To begin the argument that RTG support was actually what Commodore was planning to add to the OS, the real thing we're lacking is Kickstart and Early Boot Screen support for RTG.
How about we figure out a way to fix graphics.library and add a Kickstart module enabling a basic RTG output (at a few predefined resolutions like 640x200/640x480) allowing Early-Boot-Screen and No Startup-Sequence boots to go straight to an RTG output.

[d4rk3lf]

This is actually pretty cool idea, imho.
At least for thinking.

It's unrealistic... but not that much unrealistic.
We (only) have to wait for some Amiga fan to become billionaire, and then just throw in some millions out of fun, for the projects like this. First to finish AAA, then to pay programmers and artist to make some cool things with it.

[Pyromania]

@d4rk3lf

Why think small? I’m hopeful some Amiga fan becomes a trillionare since that’s never been achieved by one person before. Then they will have a few billion laying around for their Amiga hobby.

[Paul_s]

If AAA is ever released Medhi Ali should cut the ribbon and the cake aboard a sinking ship

[Thomas Richter]

Originally Posted by grelbfarlk

How about we figure out a way to fix graphics.library and add a Kickstart module enabling a basic RTG output (at a few predefined resolutions like 640x200/640x480) allowing Early-Boot-Screen and No Startup-Sequence boots to go straight to an RTG output.

This is not "fixing" graphics, it is "rewriting" graphics as the two big RTG implementations pretty much replace most of it. Besides, even then this would not work because you would still need to load a graphics driver for the corresponding graphics card, which is card dependent and right now sits on the disk.


However, at the time the boot menu is loaded, there is not yet any disk access, so you couldn't load an RTG driver. Thus, this cannot work, you would need some ROM-resident portion of a primitive graphics driver sitting on the corresponding board. Unfortunately, Amiga graphics cards do not work like this.


For the PC, this works because graphics cards come with their "VGA bios" which implements some very elementary IO functions to switch to a specific graphics mode, and to allow the Os to render text or (very simple) graphics to screen, unaccelerated.


We do have Autoconf on the Amiga, but no way of an Autoconf ROM providing an RTG driver.

[Pyromania]

Could WinUAE have it’s plugin architecture extended so other developers could add a AAA Chipset plugin as a experimental feature?

[Toni Wilen]

PicassoIV has flash ROM which is used to load picassoiv.resource so in theory it could have full driver in flash if needed..

WinUAE already has picasso96 driver in "ROM" (uaegfx.card) and Picasso96 supports memory resident drivers (it does not require physical file), for example AROS m68k uses it automatically to enable early boot in RTG mode.

AAA won't happen. There is no software and never will be.

[José]

I think AAA is much more fascinating than what some of you guys are making it. It's a window in the past of what would have been possible and if you're into Amiga specific things just imagine a 8x times faster blitter and a copper that would be able to change every pixel in a TV screen in 24bit ?
Now combine that with a version of the video toaster for AAA and the Amiga would have been king. There would have been full resolution 24 bit copper based effects for video!
Every time I read about the stories about how Commodore managed the development of the Amiga I get angry!

IIRC there was (is ?) a PDF on Dave Haynie's website with pretty in depth descriptions of the chipset capabilities.


Anyway, it obviously is not viable/worth the effort nowadays, except for the nostalgic factor ...


Anyone remembers the hidden message "We made Amiga, They f*cked it up" ? That one was from earlier on though...

[Glen M]

After listening to Dave Haynie speak while mingling with the crowd at Amiga Ireland I can categorically tell you this will not never ever happen for the basic reason that the chipset while produced as a first revision doesn't work, isn't even close to working. Its full of bugs of which Dave only explained one while I was listening. Apparently the best he got out of it was to display a screen at a nice resolution, something like 1200x800 if I remember correctly.

His explaining of the 1 fault went way over my head but essentially what I remember was one of the custom chips had a fault where by it corrupted the data bus so another other access requested to it was impossible. Essentially computer said no, or something to that effect anyway. The man is fantastic to listen to but his knowledge flies so far over your head that its difficult to keep track, either that or the quantity of drink consumed had err... dampened my intellect.

[demoniac]

Originally Posted by Glen M

His explaining of the 1 fault went way over my head but essentially what I remember was one of the custom chips had a fault where by it corrupted the data bus so another other access requested to it was impossible. Essentially computer said no, or something to that effect anyway.

From your description, the data coming off the bus was unreliable or wrong.

[Glen M]

Originally Posted by demoniac

From your description, the data coming off the bus was unreliable or wrong.

Something like that anyway. Maybe it wasn't so much that it corrupted the bus but rather something like the faulty chip would lock the data bus then other writes / reads to it would be impossible.

Honestly I can't remember exactly but from what I gathered there where multiple bugs. You know the way someone takes a sharp breath and jokingly says it was a long way from finished. That was the impression I got.

[Phantasm]

I see no reason why something along the lines of the Spectrum Next could not be done for the Amiga. Since the level of technical detail that would be required to recreate AAA don't appear to exist then anything done along those lines would purely be a speculation as to what might have been done.

Instead of focusing on AAA why not think about delivering a next gen Amiga the same way the Spectrum Next has done for the Speccy. An implementation could start off in software by extending one of the existing Amiga Emulators and could then be delivered in FPGA or via a dedicated emulation box (although I guess ROM copyright could be an issue).

[Glen M]

Originally Posted by Phantasm

I see no reason why something along the lines of the Spectrum Next could not be done for the Amiga. Since the level of technical detail that would be required to recreate AAA don't appear to exist then anything done along those lines would purely be a speculation as to what might have been done.

This is exactly the sort of thing the Spectrum Next is delivering. An implementation of what could have been seen as the next gen Spectrum. This would probably start off as a software implementation of a next gen Amiga and could then be delivered in FPGA or via a dedicated emulation box.

Is that not in some way what the Vampire is trying to do?

[Phantasm]

Originally Posted by Glen M

Is that not in some way what the Vampire is trying to do?

I guess so in a way - would be nice if there were a software implementation so that might encourage people to develop for it.

[Thomas Richter]

Originally Posted by Phantasm

I guess so in a way - would be nice if there were a software implementation so that might encourage people to develop for it.

The basis for the Vampire RTG graphics is P96. There is sufficient software and documentation around to support this kind of interface.