Hello Amigans,

how are you?

Is it normal that we have known each other for 20-25 years and we continue to post on these forums, waiting for the return of Amiga?

P.S.
Dudes, it's time to join the forces #??

_________________
retired

@AmigaBlitter

Quote:

time to join the forces #??

Sorry, I lost track of what # in the "join the forces" saga we are on years ago.

#6

_________________
This posting, in its entirety, represents solely the perspective of the author.
*Secrecy has served us so well*

People just work on what they like.
The community drives the platform and those that do things beyond being caught in a groove on forums are doing the things they enjoy.
A lot of development work is shared between platforms already. Often developers work on all derivatives.
Every single amiga based system has benefitted from every other many times over.

There's simply no joining forces to be had beyond what already exists.

@AmigaBlitter

Many of us work together since years if not decades.

What is your point? This is not meant negative, just
To mention.

The issue is more that some who work together in group a
Do not like (and actually hate) group b ;) i do not even
Make myselves an exception here. Though i try to break
the pattern.

@MagicSN

You might have been gone at the time, but there were about 7? of these:

It's time to join the forces!

another from back when he still numbered them

#6

Last edited by number6 on 03-May-2025 at 04:57 PM.

_________________
This posting, in its entirety, represents solely the perspective of the author.
*Secrecy has served us so well*

@AmigaBlitter

Quote:

AmigaBlitter wrote: Hello Amigans, how are you? Is it normal that we have known each other for 20-25 years and we continue to post on these forums, waiting for the return of Amiga?

How have the Jews been waiting for their messiah for three thousand years and the Christians for Jesus for two thousand years?

Well, actually, at least the Amiga might have a chance...
Quote:

P.S. Dudes, it's time to join the forces #??

That's how I do:

s = "Dudes, it's time to join the forces #??"
print("".join(s[:s.find("f") - 3].rsplit(" ", maxsplit=1)))

I really miss the Amiga Platform.

As far as i know there are many investor ready to jump in the Amiga Platform.... if only ...

_________________
retired

@AmigaBlitter who are those investors? Do you have a list?

@AmigaBlitter

Quote:

Is it normal that we have known each other for 20-25 years and we continue to post on these forums, waiting for the return of Amiga?

I think for me it was the only real "Personal Computer" I have loved using, it's annoying PC kinda refers to Windows systems, as Windows really isnt a Personal Computer, it feels very corporate and controlling and hard to understand the behind the scene and customisation is pretty poor, at least until 2000s, and even then it always feels like your fighting the computer to work maybe the way you want it to work. Never have I had that feeling on the Amiga.

Then their the community behind the scene, I always felt the Amiga scene is just more fun than most of the other communities and more nuts! But in a special kinda of way.

I wish AmigaOS5 could come out for the top end ARM chip, with a FPGA adaptor you add to have full 100% backward compatibility, had the latest big apps ported to it and offer a true alternative to Apple / Windows. Its a dream that unlikely to happen, but it fun seeing efforts and generally really interesting solutions by such a small community to push the platform forward. For that reason I will stay around and see what next for the crazy platform on the planet!

_________________
AmigaNG, YouTube, LeaveReality Studio

@AmigaBlitter

Quote:

AmigaBlitter wrote: Hello Amigans, how are you? Is it normal that we have known each other for 20-25 years and we continue to post on these forums, waiting for the return of Amiga?

Hi. I'm fine, thanks for asking.

Though of course I could be a lot better if I didn't have to use 20th century legacy computing systems and paradigms still very much present in today's versions of macOS, Windows, and Linux.

Two decades of frequenting AW.net, of good times, of awesome exchanges of ideas, and of some good chuckles. Decades of fighting with ignorance, of dashing hopes and crushing dreams. I don't regret a single minute of time I spent reading, or writing on AW.net.

I hope I'll still be showing ppcamiga1 the error of his ways for at least a couple decades more.

_________________
All the way, with 68k

It is sad.
So much time was wasted by scumbags attacking ppc on attacks on ppc Amiga Os 4 and HP.
And mui one and only thing from Amiga software that should be preserved after move to x86/arm is still not done.

Amiga still inspiring ...

https://www.hackster.io/news/linus-akesson-s-kaleidoscopico-turns-a-raspberry-pi-pico-2-into-a-powerful-demoscene-machine-85b9ba713e13?fbclid=IwY2xjawKGxkhleHRuA2FlbQIxMABicmlkETBVbmdNT0loQ3BqMHJESHR2AR5uwhakaeOdfCl9SrvS3vzMvCzhYCBNkNeyyhCsNYcafm4i1WrXO64ShiZWcA_aem_8t7p0wrxluUHhIb1xHWJ_w

Still space for an Amiga return?

_________________
retired

@AmigaBlitter

Quote:

AmigaBlitter wrote: Amiga still inspiring ... .... Still space for an Amiga return?

To throw a spanner in your works ... it will surely stop to inspire when the original Amiga users have faded away.
Most of the current Amiga scene is about gaming, and retro-gaming specifically.
And the other faction uses Amiga-like platforms for productivity purposes, or as an end in itself.
But to keep it in perspective, the market is flooded with much more impressive games on cheaper platforms. If not for nostalgic reasons, hardly anyone switches over retro gaming.
And the "productivity" niche fights an uphill battle with less performant and significantly more expensive hardware, as they must compete with current PCs and Macs for new users.

Not that I like it, but it is as it is ...

amigang Quote:

I wish AmigaOS5 could come out for the top end ARM chip, with a FPGA adaptor you add to have full 100% backward compatibility, had the latest big apps ported to it and offer a true alternative to Apple / Windows. Its a dream that unlikely to happen, but it fun seeing efforts and generally really interesting solutions by such a small community to push the platform forward. For that reason I will stay around and see what next for the crazy platform on the planet!

"100% backward compatibility" is not possible with an ARM CPU. Not even with a FPGA chipset. A 68k Amiga is made for the CPU and chipset to work together. ARM is foreign to the Amiga. Something as simple as a library open is difficult for ARM.

Open:
addq.w #1,(LIB_OPENCNT,a6)
bclr.b #LIBB_DELEXP,(LIB_FLAGS,a6)
move.l a6,d0
rts

This is short and very readable code without comments but a problem for ARM. The ARM code equivalent needs more than twice as many instructions. It is a RISC thing. PPC needs 4 instructions in a loop to replace the addq.w #1,(LIB_OPENCNT,a6) above (for 100% compatibility every atomic 68k RMW instruction like this needs to be replaced with a 4 instruction Load-Acquire and Store-Release loop and it may not be possible in some cases due to alignment restrictions). A 5th PPC isync or similar instruction is likely required after this too. ARM64/AArch64 is likely to be similar. The original ARM ISA did not support 16-bit load/stores and the original Alpha ISA did not support 16-bit or 8-bit load/stores so several more instructions would be necessary on them. There is a RISC performance handicap to go with the RISC code bloat too.

codis Quote:

To throw a spanner in your works ... it will surely stop to inspire when the original Amiga users have faded away. Most of the current Amiga scene is about gaming, and retro-gaming specifically. And the other faction uses Amiga-like platforms for productivity purposes, or as an end in itself. But to keep it in perspective, the market is flooded with much more impressive games on cheaper platforms. If not for nostalgic reasons, hardly anyone switches over retro gaming. And the "productivity" niche fights an uphill battle with less performant and significantly more expensive hardware, as they must compete with current PCs and Macs for new users. Not that I like it, but it is as it is ...

Did you read the last article AmigaBlitter linked? He is talking about low end Amiga like hardware. People are still trying to recreate Amiga like hardware after all these years. "the Raspberry Pi RP2350-powered Raspberry Pi Pico 2" does not have video or sound output so it was created. One of the two CPU cores was used to create the equivalent of the 68k Amiga custom chips. The end result was more difficult to create than a 68k Amiga demo and I have seen better demos and games on 1985 68k Amiga hardware. The RP2350 SoC ASIC is about $1 USD and uses more transistors than a 68060 and AA+ Amiga chipset. A 68060&AA+ SoC ASIC Amiga with HDMI output could be mass produced for a similar price and use AmigaOS where standard small footprint OSs are lacking. There are all kinds of RP2350 and RPi Pico projects and articles and millions of these MCUs sold where a 68k Amiga could do more better. This would have been easily possible with the finances wasted on lawsuits and bloated "desktop" PPC hardware where the AmigaOS lacks desktop features. The closed and expensive Amiga like hardware has kept the Amiga from proliferating and developing like it should. I think AmigaBlitter gets it. Do you?

Last edited by matthey on 06-May-2025 at 09:29 PM.

@AmigaBlitter

Hey welcome back!

I have focused primarily on MorphOS these days.

It has been actively developed over the last 5 years and has really nice productivity software, including a modern browser and email client.

The best system to run it on is still a PCIe PowerMac G5 @2.5GHz. Drives 2 monitors well and is a joy to use. They released 3.19 a few months ago.

So all in your out a few hundred bucks for a nice experience.

I daily drive mine for work and enjoyed the drive...

@matthey

Quote:

"100% backward compatibility" is not possible with an ARM CPU. Not even with a FPGA chipset.

You must also add “Not even a real Amiga.�

My FPGA systems are heads above my “real� Amigas when it comes to compatibility with old software. Real AGA often struggles to be compatible with software written for OCS/ECS and vice versa naturally is a no go, and anything beyond real 68030 provides tons of compatibility problems as well, real 68030/040/060 existing in half a dozen variants (added up) doesn’t help. In contrast, FPGA systems can switch between chipsets, cpu and RAM configurations on the fly, or at worst a reboot.

Last edited by kolla on 07-May-2025 at 01:05 AM.

_________________
B5D6A1D019D5D45BCC56F4782AC220D8B3E2A6CC

@kolla
The problem is not FPGAs but emulation. A FPGA can have the exact logic and timing of the original chip up to a limit. Much parallel logic is possible in a FPGA per clock cycle but not as much sequential logic. Sequential logic is more important for CPUs so the limit is relatively low while chipsets and GPUs often have simpler parallel logic. The field programmable ability of FPGAs gives tremendous versatility. Still, an ASIC is just as impressive if not more impressive in other areas like transistors/$, performance, performance/$, performance/W, customization, etc. Emulation is programmable and the easiest to create but the least accurate recreation of logic, especially without executing on a RTOS. Emulation may be useful for rough simulations of hardware while FPGAs can provide exact simulations of low clocked hardware. ASICs are the final destination where high clock speeds, many transistors, low power and mass production cost are important.

Some of the incompatibility is easier to solve today. The AGA chip mem fetch alignment requirement can likely be done away with for example. Actually, if Jay Miner's solution of dual ported VRAM in the Ranger chipset had been used then this limitation would not exist. SRAM blocks in a FPGA can likely be dual ported too for chip mem but they are limited to a small size and other solutions with large bandwidth external memory may also be able to remove the alignment requirement. VRAM is no longer available but an ASIC could potentially use dual ported eDRAM for large amounts of dual ported memory which is not possible with a FPGA. An ASIC could allow more practical CPU MMU use without as much performance loss. A MMU can improve compatibility with memory remapping, tricks like bug fixes and could even help to implement 68k SMP with an idea I have (shared data pages would use locked RMW memory accesses). It is also possible to license eFPGA blocks for an ASIC. FPGAs could be thought of as ASICs with eFPGA blocks and other custom hard blocks like memory controllers, DSPs and even CPU cores. The only problem is that the CPU cores are not the architectures we need for retro use.

Last edited by matthey on 07-May-2025 at 04:28 PM.
Last edited by matthey on 07-May-2025 at 04:26 PM.

@amigang

Quote:

amigang wrote: @AmigaBlitter I wish AmigaOS5 could come out for the top end ARM chip, with a FPGA adaptor you add to have full 100% backward compatibility,

An FPGA can't have 100% compatibility without having the original components (processor and chipset) perfectly reimplemented.

Which usually requires the original schematics.


@kolla

Quote:

kolla wrote: @matthey Quote: "100% backward compatibility" is not possible with an ARM CPU. Not even with a FPGA chipset. You must also add “Not even a real Amiga.� My FPGA systems are heads above my “real� Amigas when it comes to compatibility with old software. Real AGA often struggles to be compatible with software written for OCS/ECS and vice versa naturally is a no go, and anything beyond real 68030 provides tons of compatibility problems as well, real 68030/040/060 existing in half a dozen variants (added up) doesn’t help. In contrast, FPGA systems can switch between chipsets, cpu and RAM configurations on the fly, or at worst a reboot.

Same as above: no perfect chips implementation, no party.

@AmigaBlitter

Quote:

AmigaBlitter wrote: Amiga still inspiring ... https://www.hackster.io/news/linus-akesson-s-kaleidoscopico-turns-a-raspberry-pi-pico-2-into-a-powerful-demoscene-machine-85b9ba713e13?fbclid=IwY2xjawKGxkhleHRuA2FlbQIxMABicmlkETBVbmdNT0loQ3BqMHJESHR2AR5uwhakaeOdfCl9SrvS3vzMvCzhYCBNkNeyyhCsNYcafm4i1WrXO64ShiZWcA_aem_8t7p0wrxluUHhIb1xHWJ_w Still space for an Amiga return?

I assume that you haven't read the full news and, even more important, project's page, right?

Because it tries to emulate the Amiga systems in some ways, but it's a far away imitation.

It's wasting so many resources to achieve things which are super efficient and much, much better done on an Amiga, and it doesn't even resemble the philosophy of our beloved machines.

The same resources could have been better used to create an Amiga successor which is far above both AAA and Hombre (processors included).

@matthey

Quote:

matthey wrote: amigang Quote: I wish AmigaOS5 could come out for the top end ARM chip, with a FPGA adaptor you add to have full 100% backward compatibility, had the latest big apps ported to it and offer a true alternative to Apple / Windows. Its a dream that unlikely to happen, but it fun seeing efforts and generally really interesting solutions by such a small community to push the platform forward. For that reason I will stay around and see what next for the crazy platform on the planet! "100% backward compatibility" is not possible with an ARM CPU. Not even with a FPGA chipset. A 68k Amiga is made for the CPU and chipset to work together. ARM is foreign to the Amiga. Something as simple as a library open is difficult for ARM.

It's possible both on software and FPGA side, but provided the schematics to perfectly reproduce the original chips.

WinUAE seems to have a cycle-accurate 68000 emulation, and the upcoming version has a very accurate chipset emulation thanks to the use of some (unfortunately not all) schematics found for Alice.

But nothing prevents to have a perfect emulation of all other components (included a 68EC020), getting their schematics.
Quote:

Open: addq.w #1,(LIB_OPENCNT,a6) bclr.b #LIBB_DELEXP,(LIB_FLAGS,a6) move.l a6,d0 rts This is short and very readable code without comments but a problem for ARM. The ARM code equivalent needs more than twice as many instructions. It is a RISC thing. PPC needs 4 instructions in a loop to replace the addq.w #1,(LIB_OPENCNT,a6) above (for 100% compatibility every atomic 68k RMW instruction like this needs to be replaced with a 4 instruction Load-Acquire and Store-Release loop and it may not be possible in some cases due to alignment restrictions).

Such 68k instructions aren't atomic: the RMW cycle doesn't guarantee that the memory location is atomically updated.

Only TAS, CAS, CAS2 68k instructions are atomic, but even those can be emulated by architectures having load/store atomics (not x86/x64: this requires transaction extensions, which are a bit more complicated to be used).