AmigaOne Hardware Design

[DaveP]

Why did Amiga decide to design/develop custom hardware for the new AmigaOne platform?

They did not. It is a ( close enough ) POP design called the Teron CX with AmigaOS routines in firmware.

Since they've made a system that can run Linux or AmigaOS out-of-the-box,

So can the original Amiga with APUS or BSD3.x

 why didn't they just design the AmigaOS to run on x86-based hardware to begin with without the need for a custom-designed motherboard?

Repeat the motherboard is not custom designed. Secondly AmigaOS is hard enough to port to PPC from the mish mash of 68k asm, BCPL and C. Call it a transition. But then do you have any faith that Amiga Inc wont go tits up and leave Hyperion to run things? I dont and I hope they do. In which case AmigaOS will stay PPC in the near term.

What advantages will we have by using a custom-designed motherboard, when there aren't really any kick-@ss custom chips on it

It isnt custom.

(e.g., next-rev. Hombre).

Vapourware.

 I mean, ArticiaS, AGP 2x, no firewire support... What gives? If they'd have gone with a Linux-base and worked from there, we'd at least have drivers for gfx/sound cards... I guess my point is, if you're gonna make it a custom job, then give it some rockin' custom chips.

It does have drivers for GFX and sound cards.

eh?

You make no sense to me bud, hope that has helped you some.

Also, isn't the rumor that they will eventually make AmigaOS for x86-based machines, anyway? Why not cut out the middle-man, then, and cut out the custom design stuff?

[DaveP]

Well, thanks for the info. I guess I thought the A1-board was custom-designed since it took them an eternity to get the machine shipped.

You would think wouldnt you? :-) No it took an eternity to ship because ( this is second hand info ):

1. The original boards to be shipped in April were tested by Alan and he found that the BIOS ( firmware ) was totally buggy causing serious problems for the development team. He polled the recipients and they all agreed to get them sent back until MAI had fixed this.

2. The subcontractors that MAI called into to fix this totally ####ed up. Hyperion took over and got from zero to a working BIOS in about three months, causing a three month delay in shipping Exec SG to the AOS4 development team.

3. Alan decided he couldnt hold out any further and announced he was shipping SE boards ( fixed by Hyperion ) in December, followed by XEs in the first part of the new year.

The minor fixes that have been found to be required ( in the APRIL chip for pegasos and a few more that are not documented publically ) have been worked around and MAI are making permenant changes.

If nothing was custom, why didn't they select better components for the base system? Why couldn't they have cranked it out in a couple weeks?

For the second part see above, for the first part well that was what was available.

I guess what I'm asking is... what is it about the A1 board that makes it so special when there are tons of other board designs out there that were already superior when Amiga released the design specs for their system? Am I missing the point entirely?

It is special because it is the first Amiga board without custom chips to run AmigaOS4.

This is also a transition, first AmigaOS has to remove the custom chip dependency that was great to start with and quickly became a problem.

You say superior board designs, well IBM has a few but you need to be able to prove reasonable volume sales are available up front in order to be a vendor.

But I think you are angling at x86. Well, as I said, get the OS reasonably portable before making the leap to x86 and x86 is going through a transitional year.

Caught between Linux and Windows on the x86 AmigaOS would definately die. Caught only between Linux and "morphos" AmigaOS has a slightly better chance.