@terminills

Quote:

Honestly if it was up to me I would make sure stuff like the updated Final Writer never runs on PowerPC just out of spite because you act like a tool.

Isn't it a little unfair of you to judge the entire PowerPC community by a single forum troll?

_________________
The Rear Window blog

AmigaOne X5000/020 @ 2GHz / 4GB RAM / Radeon RX 560 / ESI Juli@ / AmigaOS 4.1 Final Edition
SAM440ep-flex @ 667MHz / 1GB RAM / Radeon 9250 / AmigaOS 4.1 Final Edition

@Karlos

https://amigaworld.net/modules/newbb/viewtopic.php?topic_id=45375&start=0&post_id=875214&order=0&viewmode=flat&pid=0&forum=47#875214

_________________
Support AROS sponsor a developer.

"AROS is prolly illegal ~ Evert Carton" intentionally quoted out of context for dramatic effect

@Trixie

Technically it's not judging but punishing and you are correct which is why I wouldn't.

_________________
Support AROS sponsor a developer.

"AROS is prolly illegal ~ Evert Carton" intentionally quoted out of context for dramatic effect

Just for the avoidance of doubt, I am not a PPC hater. I just think it has ran aground many years ago. An evolutionary dead end, so to speak. Outside of our tiny community, nobody cares and all the other software it ran is also available for other architectures already.

By comparison, 68K has transcended. We have physical and software (post physical, lol) implementations that are hugely faster. We have a sizeable back catalogue of 68K software and we still live in the architectural shadow of 68K on the PPC NG front. We didn't go 64 bit, we didn't go SMP. Does anyone really think those are going to happen now?

Seriously the best way to leverage those additional CPU resources now is by using something else, but if you want to throw the resource at Amiga compatibility, then just have a bare metal emulation.

_________________
Doing stupid things for fun...

@terminills

Quote:

terminills wrote: @Karlos https://amigaworld.net/modules/newbb/viewtopic.php?topic_id=45375&start=0&post_id=875214&order=0&viewmode=flat&pid=0&forum=47#875214

"As terrible as Zoidberg's idea is, it's brilliant when I have it!" -- Bender

_________________
Doing stupid things for fun...

@Karlos



Quote:

"As terrible as Zoidberg's idea is, it's brilliant when I have it!" -- Bender

I've had the idea for a while. :D

I'd say it's a little closer to Draco than Amithlon though as it requires patching the OS in memory(AmigaOS) vs Emulating a full system. :)

Last edited by terminills on 21-Dec-2024 at 06:46 PM.

_________________
Support AROS sponsor a developer.

"AROS is prolly illegal ~ Evert Carton" intentionally quoted out of context for dramatic effect

@terminills

Quote:

terminills wrote: I'd say it's a little closer to Draco than Amithlon though as it requires patching the OS in memory(AmigaOS) vs Emulating a full system. :)

Amithlon is x86 Draco.

_________________
All the way, with 68k

@Gebrochen

Quote:

Gebrochen wrote: Also, People saying Amiga OS 4 is dead in the water, er, well, for all what we know, they've been working hard in the background to give us our wish, Amiga OS4.X on Arm / X86 / Snapdragon, damn, I dont know, just something more common and cheaper to purchase to get out of a current stale mate niche.... I mean they did unexpectedly release Amiga OS 3.1.4 then 3.2.1, etc.... I do not recall being expecting that to ever happen

A port of AmigaOS 4 to another ISA is a greater undertaking than the improvements made to OS 3, mostly through back porting OS 4 components.
It’s a principle of Information Theory. Keeping such a large endeavor a secret costs a lot of money, which is something Hyperion do not have.

So yes, AmigaOS 4 has gone a far as it can go with Hyperion. Just as PowerPC has gone as far as it can go with AmigaOS and MorphOS systems. Which is why we are seeing so much activity in emulating 68k/Amiga OS 3 on ARM, and the increased interest in AROS.

_________________
All the way, with 68k

@agami

Not entirely.

_________________
Support AROS sponsor a developer.

"AROS is prolly illegal ~ Evert Carton" intentionally quoted out of context for dramatic effect

Ahead-of-time (AOT) compilation makes more sense than JIT compilation for 68k Amiga execution on LE hardware. BE mode ARM JIT is far from future proof with BE ARM support being minimized. A BE 68k intermediate language/code does not make sense for LE hardware. There is sometimes swizzling overhead even for AOT compilation which is an improvement over JIT. A native LE AmigaOS makes the most sense from a technical perspective but further distances a modernized Amiga from the original retro 68k Amiga market making it a bad idea from a marketing perspective. The BE PPC AmigaOS was already so incompatible with 68k Amiga software and so far removed from what the 68k Amiga was that it only appealed to a tiny percentage of Amiga users. The virtual LE Amiga Anywhere was worse as it was not only Amiga Nowhere from low adoption rates but also from not having anything to do with the 68k Amiga.

A virtual Amiga limits appeal outside of the retro 68k Amiga market. There is no appeal to other 68k markets where other retro 68k markets are likely larger than the 68k Amiga market. Emulation is noncompetitive for general purpose and embedded use where a virtual 68k Amiga will be inferior in every CPU metric compared to real hardware. Half of RPi sales are for embedded use and the CD32 SBC was used for embedded use but nobody is using THEA500 Mini, A600GS, (Win)UAE, Emu68/PiStorm or AC/Vamp "Amigas" for embedded use and never will (high performance FPGA CPU cores are not competitive either). Modernized Amiga small footprint standard hardware still has advantages with real low end hardware but the hardware and the AmigaOS need to be very cheap. The AmigaOS is practically worthless for desktop use and low value for low end retro, hobby and embedded use and needs to be cheap. Trying to survive off expensive AmigaOS profits against free alternatives that are better featured and more mature is brain dead. I believe the only chance for the Amiga is to produce competitive 68k Amiga/universal hardware again and compete with RPi like hardware and toys. Developing 68k CPU cores is more difficult and has higher up front costs than using ARM cores but ARM is not a giant to be feared. It is possible to unite practically all Amiga users and bring back old Amiga fans with modernized 68k Amiga hardware. There is so much discussion about porting the AmigaOS to other architectures but all ports do is divide the Amiga user base with hardware that is less popular than the 68k Amiga. An AmigaOS port would be late to x86-64 and perhaps ARM where a more mature AROS and MorphOS would compete against it. AROS x86-64 uses some of the most powerful hardware, supports 64-bit and has the best SMP support for an Amiga like OS but is going nowhere. The grass is greener on our 68k Amiga side of the fence.

@amigakit

Looks like a blue A1200 board.

@Yssing

Quote:

I am excited about the potential move to ARM over PPC. We need cheaper hardware in order to get more users.

But, ARM is still not 68K, so would this fail? I notice there is a lack of ARM coprocessors on Amiga cards used in the same way as the PPC cards. The ones that feature an ARM lack a matching kernel to easily call ARM code blocks from 68K side. Perhaps there is a reluctance to cloud the Amiga with another CPU to pick up the slack from 68K that occurred with PPC. Popular uses like PiStorm avoid native ARM hooks and stick to emulating 68K fast on ARM.

@Karlos

Quote:

This is why, for a long time, I've been of the opinion that in virtuo 68K is the way to go. We abstract out everything else in our need to support as much hardware in as sane a way as possible, so why not the CPU?

What I wonder is is a software solution would work as well? What I mean is being able recompile OS4 sources into different CPU targets. Similar to to what PortableE does.

There's this language called EEL. It's like a cut down C and looks similar. They have this compiler that takes sources and compiles it on the fly into native code. Suppose it's similar to JavaScript JIT in a way and close to WASM idea but more high level.

They actively supported PPC until a few years back. Support x86/64 and recently ARM. Being able to take one source and run it on any target is real useful. Heck, use it for scripting, and it could knock Arexx's socks off!

https://github.com/justinfrankel/WDL/tree/main/WDL/eel2

@matthey

No one using an Amiga or Clone really cares. PPC is dead, technically 68K is dead even though it's quite active for Amiga standards. but at the end of the day most just want to use the programs they want and enjoy themselves.

_________________
Support AROS sponsor a developer.

"AROS is prolly illegal ~ Evert Carton" intentionally quoted out of context for dramatic effect

What I hope it is:

1. New A1200 motherboard with real chips or...

2. New A1200 motherboard with FPGA



What I think it is (but I am not interested in):

1. Some Amiga NG board in the shape of an A1200 motherboard to fit an original A1200 case or...

2. Raspberry Pi in the shape of an A1200 motherboard to fit an original A1200 case

Last edited by Everblue on 22-Dec-2024 at 08:36 AM.

_________________

@Hypex

A software solution like that is a solution to a slightly different problem. My preference for a 68K runtime is in part because of the large catalogue of older software that only exists as binaries, that still run today either under emulation or on actual silicon.

_________________
Doing stupid things for fun...

@matthey

Quote:

Ahead-of-time (AOT) compilation makes more sense than JIT compilation for 68k Amiga execution on LE hardware.Â

Not necessarily. Not all code is amenable to being fully translated ahead of time and I suspect that under AmigaOS/68K the corpus of code that would actually work well here is small. Which then defeats the whole point of the endeavour.

Of course, you can have both mechanisms. However, in fretting over JIT you are consistently overlooking an important benefit: JIT compilation and hot path optimisation can produce native code that's fine tuned for your exact configuration at the time you run it, taking into consideration many factors that no AOT compilation can reliably try to predict.

Emu68 and UAE both utterly shatter the performance of all extant 68K silicon (your ASIC is not real and likely never will be) and for something intended to be backward compatible with software that runs on extant silicon at tens of MHz, you are worried about the performance overhead of said emulation?

Just go back and look at the Lightwave tests. Some of the most arithmetically dense, unpredictably branchy, scattered memory access workload imaginable (that's raytracing). UAE and Emu68 both ate it for breakfast and asked for more.

Honestly, I don't get the objection. Sure I get that for some people only physical silicon is real and they'll never settle for anything less. That's fine, there are plenty of accelerator solutions available today. Nobody is taking anything from them by providing an alternative that can run on commodity hardware.

_________________
Doing stupid things for fun...

@Karlos

Actually running modern hardware is important, major different between DDR1, DDR2, DDR3, DDR4 and DDR5, AmigaONE PowerPC is currently on DDR3 at best, not sure what QorIQ Txxx cpu’s supports but, if your not keeping up, the CPU that are, will be able to emulate your CPU’s.. I’m not sure how fast memory can get, or if we are closing the end of the hardware race.

Sure, we want the latest PC games to run on AmigaOS, but its not going to happen without SMP

Last edited by NutsAboutAmiga on 22-Dec-2024 at 12:22 PM.
Last edited by NutsAboutAmiga on 22-Dec-2024 at 12:19 PM.

_________________
http://lifeofliveforit.blogspot.no/
Facebook::LiveForIt Software for AmigaOS

@NutsAboutAmiga

SMP is a problem space that if it had an easy, backwards comaptible solution, would have been found ages ago. I don't believe it was for a lack of desire and effort. Whatever anyone says about the PPC/NG proposition, SMP capable hardware was made available, what, 12 years ago?

However, one thing to consider about an in-virtuo solution is that you can implement something that never existed: Want a dual 68K configuration? I am sure that's not an outlandish proposition for a bare metal emulator on natively multicore hardware. I am equally sure there will be some complications over the existing single processor model. However, there is also tantalising possibility here that the complications that go with "real SMP" on "real Hardware" can be worked around or met half-way, since you are in charge of how your virtual dual core 68K actually works, so issues like affinity and cache coherency are yours to dictate.

Last edited by Karlos on 22-Dec-2024 at 12:35 PM.

_________________
Doing stupid things for fun...

@Karlos

Quote:

Karlos wrote: @NutsAboutAmiga SMP is a problem space that if it had an easy, backwards comaptible solution, would have been found ages ago. I don't believe it was for a lack of desire and effort.

SMP and 64Bit are both nogos inside the current API/ABI and while half a#### attempts may kinda work they would still be a bad decision.

Doing as a non EXEC "compute.library" similar to PowerUP or the x86 modules in Amithlon should be feasible (and is already kinda been done in some emulation solutions).

Add a proper API that still looks "Amiga" to the point where one can have proper apps running on the dark side (calling back to "Amiga" for GUI and such) and all would be good.

Chicken and Egg plus Not Invented Here syndromes would IMO be a bigger obstacle as the actual implementation as we would be talking about yet another AmigaNG option competing and not cooperating with the existing one, while being far to short on manpower/userbase to really pull it of.

_________________
- We don't need good ideas, we haven't run out on bad ones yet
- blame Canada