Linked by Thom Holwerda on Thu 26th Mar 2009 23:34 UTC
Amiga & AROS Despite the recent emergence of several new ways to actually run AmigaOS 4.0, the supply of machines is still extremely small, and not very future proof. As such, one of the most recurring questions within the Amiga community is why don't they port the darn thing to x86?
Thread beginning with comment 355311
To read all comments associated with this story, please click here.
Is it special under the hood?
by John Blink on Fri 27th Mar 2009 01:53 UTC
John Blink
Member since:
2005-10-11

Or is it the interface and programs that most people desire out of the AmigaOS?

I would really like to know.

Reply Score: 1

SamuraiCrow Member since:
2005-11-19

Or is it the interface and programs that most people desire out of the AmigaOS?

I would really like to know.


It's the responsiveness of the OS that makes it popular. And, years ago, the custom graphics accelerator chips made it very well integrated. (Read that: No drivers necessary since the the hardware came from the same company as the operating system.)

Although I have a MicroA1-c that runs AmigaOS 4.1, the technology I'm watching is the Natami which promises to have an improved Amiga compatible multimedia chip set and 68030 compatible processor all in an FPGA. My only complaint about that design is that it will be running AmigaOS 3.9 since 4.0+ requires a PowerPC accelerator board.

@thread

As far as I'm concerned about AmigaOS on Intel architecture, Intel can go fly a kite. Multicore processor support isn't present in AmigaOS so that's out of the question. Intel used this wierd idea called "little endianness" which will reduce the performance of AmigaOS to the current PowerPC performance levels anyway due to all the BSWAP opcodes required to emulate big endianness on an x86.

If Hyperion Entertainment VOF was going to port to any new architecture at all, I'd recommend LLVM. It runs on either x86, PowerPC, or ARM with roughly equal efficiency. It's open source so any new architecture that comes out can be supported without requiring the entire source code of the OS (assuming they distribute it as LLVM bitcode-format rather than straight binary).

Reply Parent Score: 4

galvanash Member since:
2006-01-25

Intel used this wierd idea called "little endianness" which will reduce the performance of AmigaOS to the current PowerPC performance levels anyway due to all the BSWAP opcodes required to emulate big endianness on an x86.


Considering that the Intel 8080 is widely considered the first usable general purpose microprocessor, and that it and nearly all of the early CPUs (VAX, MOS 6502, Z80) with the exception of the 6800 were little endian - I find it odd you call it a "weird idea". PowerPC isn't big endian anyway, it can do either. There are pros and cons to both approaches imo.

Regardless, any competent port of Amiga OS to x86 would I would hope switch the implementation to little endian to avoid the issue entirely. It might make porting of some low level code a bit more difficult than it otherwise would be, but it would be better in the long run imo.

But still, if you think the comparatively minor overhead of byte order swapping would in any way negate the HUGE performance advantage modern x86 hardware has over the dinky PowerPC hardware currently available for AmigaOS your nuts.

Reply Parent Score: 4

FunkyELF Member since:
2006-07-26

If Hyperion Entertainment VOF was going to port to any new architecture at all, I'd recommend LLVM. It runs on either x86, PowerPC, or ARM with roughly equal efficiency.


Interesting. Are there currently any LLVM operating systems out there?

Reply Parent Score: 2

bugjacobs Member since:
2009-01-03

Its so much nostalgia for what it WAS back then,
with its for the time so unique custom chips, pretty good if not the best.

It was the last of the "home computers" before the split between consoles and PCs finally happened for the mainstream.

And with the demise of the Amiga the large demoscene went in a steady decline, never to be the same - thats atleast how it feels for me, the demoscene is still there, but the time when virtually ANY geek could hack on the bare metal of a computer and understand it fully from top to bottom must have ended back then.

Now computers are too complex for the average joe to start hacking. Sure anyone can get into webscripting and such, and there are plenty of BASIC compilers out there , but its not the same.

With the C64 and the Amiga virtually anyone could have a total image of the computer in their conscious mind, its capabilities and how to utilize them , now thats left to experts IMHO and you need these high level programming languages for everything, gone are the days of "banging the metal" (or silicon that is) .. :-)

Amiga for me was the tightknit custom chips and the OS and the SCENE around it. The OS was just a part of it.

I really hope the NATAMI guys manage to relase something btw. www.natami.net , They are recreating an Amiga in an FPGA chip, and this Amiga has additional features ! A product targeted for anyone wanting to revive the demoscene and do some Amiga Retrocomputing the way it could have been as the next step for the legacy .. ! SuperAGA and a highpowered 68000 class CPU core.

Reply Parent Score: 2

jal_ Member since:
2006-11-02

I really hope the NATAMI guys manage to release something


Probably not, the last message is from October (not counting the server move) saying something that "soon!" there will be hardware. That's almost half a year ago.

Sadly, these projects seem always doomed to fail. Jens S. and his Clone-A also seem to have disappeared (well, the Clone-A at least, I saw Jens last year at Breakpoint, but no mention from Clone-A, which he so proudly presented the year before) - there's also no mention on his site.


JAL

Reply Parent Score: 1