@BigD

Quote:

Are you for real?! 50 measly Euro!!! And that is why no sane developer never mind greedy Hyperion lawyers would invest so much in a product that you would only spend 50 Euros on! Heck AmigaOS 3.2 without Kickstart Roms or Floppys costs, wait for it... ... approximately 44.11 Euro!!! And that is based on a hell of a lot of legacy 68k code! The Arm code would have to be written from scratch! Who would you contract to do the work for your 50 Euro x 1000 other RPi AmigaOS paid up users (let's be optimistic)? That's 50,000 Euro for what? 5 years work to get basic functionality without even a JIT PPC emulator? 10,000 Euro a year to a solo developer? Is that what you envisage?

In my personal and extremely humble opinion, and despite the dramatic lack of modern software availability, the value proposition of a native Raspberry Pi port of AmigaOS 4.x would be FAR more interesting than any legacy 68K upgrades, generally available for 25+ years old hardware.

As far as I understand 1000 is more or less the magnitude of the current OS4.x userbase. I'm quite confident that with a cheap, reasonably powerful and definitely not outdated HW like the RPi the platform might gauge a way bigger interest, but that's just my opinion. I know I would bite though, that's for sure.

Please note, as it's already been stated, OS4 is already for the most part written in C so while definitely not a trivial job, I believe such a port it could be easier (hence cheaper) than you think.

Last edited by Turrican3 on 26-May-2021 at 03:49 PM.

@thread

What do you think is a good number maybe the Magic Number for a port?
I was wondering if 50,000 U.S. Dollars would be a good number.

50000 ÷ 131 = 381.69 U.S Dollars each.

_________________
John 3:16

Turrican3 Quote:

In my personal and extremely humble opinion, and despite the dramatic lack of modern software availability, the value proposition of a native Raspberry Pi port of AmigaOS 4.x would be FAR more interesting than any legacy 68K upgrades, generally available for 25+ years old hardware.

RPi quick BE AmigaOS 4 port
+ cheap to port
+ AmigaOS source code compatibility
+ faster 68k and PPC emulation
- outdated desktop features like no SMP, 64 bit addressing, memory protection and resource tracking
- late to market
- hardware with BE support may disappear
- only appeals to diehard Amiga fans

RPi proper LE AmigaOS 4 port with upgrades for SMP, 64 bit, memory protection and resource tracking
+ modern OS features like SMP, 64 bit addressing, memory protection and resource tracking
+ easier porting of LE x86-64 and ARM software
? may appeal as a lightweight OS but may not be lightweight anymore after adding modern features
- more expensive than current AmigaOS 4 to cover major development costs (competition is free)
- poor AmigaOS compatibility as the AmigaOS API is redesigned and LE
- slow 68k and PPC Amiga emulation
- very late to market

The first porting option above only appeals to diehard Amiga fans. The second porting option would be more appealing to RPi users if there was software available and it was free but neither is likely. Furthermore, it would be difficult to get RPi users to try a new OS even if it was free and they were payed to try it. It is very difficult to successfully introduce a new desktop OS today, especially without being installed on new hardware. Many large companies with good reputations and large budgets have failed.

AmigaOS appeal on RPi
+ diehard Amiga users
- ex-Amiga users
- non-Amiga users

Low end hardware for the 68k Amiga could be made cheap if mass produced. There is the potential to appeal to embedded markets and toy/hobbyist 68k markets (68k gaming) expanding the Amiga market. Embedded markets are what made ARM hardware cheap due to mass production but this advantage can be leveraged by other architectures which are appealing for embedded use. Any transistor overhead of 68k cores are more than made up by a savings in caches and memory. On die transistors are cheap anyway.

AmigaOS appeal on cheap 68k hardware
+ diehard Amiga users
+ ex-Amiga users
+ non-Amiga users (for embedded and 68k gaming)

Turrican3 Quote:

As far as I understand 1000 is more or less the magnitude of the current OS4.x userbase. I'm quite confident that with a cheap, reasonably powerful and definitely not outdated HW like the RPi the platform might gauge a way bigger interest, but that's just my opinion. I know I would bite though, that's for sure.

Cheap hardware with an expensive outdated desktop OS which costs several times what the hardware cost is not appealing except to a few diehard Amiga users. It is the whole package of hardware and software which needs to be appealing.

Turrican3 Quote:

Please note, as it's already been stated, OS4 is already for the most part written in C so while definitely not a trivial job, I believe such a port it could be easier (hence cheaper) than you think.

The poor Hyperion AmigaOS 4 port to ARM probably wouldn't be too expensive but a modern desktop AmigaOS NG port to ARM would likely be very expensive with little hope to recoup costs.

@matthey

Quote:

Cheap hardware with an expensive outdated desktop OS which costs several times what the hardware cost is not appealing except to a few diehard Amiga users. It is the whole package of hardware and software which needs to be appealing.

Uh? It would not be "several times" as expensive, at least not in the scenario I'm proposing. To tell you the truth I've had a look at my archive and I paid something like 25€ (+shipping) for AmigaOS 4.1 Classic.

Having said that, my point is that OS 4.x on RPi would be more appealing compared to the current scenario, i.e. where compatibile PPC hardware is roughly 10x as expensive minimum if I'm not mistaken.

Having the OS ported to a 1/10th cheaper device would definitely make the potential interested audience way bigger than it currently is IMHO, and hopefully wouldn't be too expensive. A true NG AmigaOS might come later, I can certainly agree it wouldn't make financial sense starting with that unless we magically find a huge Amiga investor with deep pockets.


I also disagree about appeal of AmigaOS on cheap 68K hardware.

Ex-Amiga users left the platform for a reason; they're not coming back for a few additions to a 68K system that was first and foremost about the games. And regarding the latter, I'd argue that AmigaOS is not a strict requirement unless you want to go the WHDLoad route (and even then, you definitely don't need new features)

Last edited by Turrican3 on 27-May-2021 at 01:07 PM.

@Turrican3

Why should new features and investment be put into AmigaOS NG if AmigaOS Classic has no value? Neither can be used as a daily driver in their current form but there are enough cheap enough ways of getting back into the 68k Classic Scene to make it a fun and fairly inexpensive hobby! The Tabor was attempting to make the same true for AmigaOS 4.x but it hasn't transpired! If there is not enough support for the belated Tabor and Hyperion are losing interest in AmigaOS 4.x then AmigaOS Classic WILL be the way forward because that will be the ONLY active community the way things are going!

In regards to the ex-Amigans leaving for a reason etc. I don't think that they're aware of how fun and active the Amiga scene still is on Classic! A lot of them still have their machines in their attics and I've actively tried to encourage my friend to show his children the machine only recently! I don't think you'd be able to encourage them in quite the same way with an AmigaONE and I DO think they have value I just can't see that the market can be grown like the Classic one can be!

Last edited by BigD on 27-May-2021 at 03:42 PM.

_________________
"Art challenges technology. Technology inspires the art."
John Lasseter, Co-Founder of Pixar Animation Studios

@Trekiej

135 Voters.

_________________
John 3:16

@BigD

Quote:

Why should new features and investment be put into AmigaOS NG if AmigaOS Classic has no value?

Because I'm assuming 4.x is the branch that has a higher chance of surviving. And since it's the more up to date code branch, i'm 99.9% sure it needs less work to be ported. Plus, I'm calling for the bare OS to be ported first, then assuming a certain degree of success, you might be able to plan a bigger investment that adds new features.

I can certainly agree that a Classic Amiga (but not necessarily its OS, which I still love by the way) can look more interesting to the vast majority of the old users. But I don't see this as an issue, since it's far more easier to go the emulation route - perhaps with a dedicated frontend, which is what I'm expecting for the THEA500 - instead of relying on 25+ years old hardware that might fail any day and requires a fairly ancient TV/monitor set. Not to mention it's often not that cheaper to purchase nowadays.

This is why I believe ClassicOS is a dead end and a 4.x port to modern, cheap standard hardware might be a far better solution if (and that's a big if, sure) we want a future for both AmigaOS and its legacy software lineup, i.e. games.

Last edited by Turrican3 on 27-May-2021 at 10:40 PM.
Last edited by Turrican3 on 27-May-2021 at 09:57 PM.

Turrican3 Quote:

Uh? It would not be "several times" as expensive, at least not in the scenario I'm proposing. To tell you the truth I've had a look at my archive and I paid something like 25€ (+shipping) for AmigaOS 4.1 Classic.

The 25€ price for AmigaOS 4.1 is the stopped development price. It used to be 99€ while being developed which is about three times the price of a RPi. Hyperion was likely losing money on AmigaOS 4 at 99€ which is why they stopped development so the price needed to be higher to break even.

Turrican3 Quote:

Having said that, my point is that OS 4.x on RPi would be more appealing compared to the current scenario, i.e. where compatibile PPC hardware is roughly 10x as expensive minimum if I'm not mistaken. Having the OS ported to a 1/10th cheaper device would definitely make the potential interested audience way bigger than it currently is IMHO, and hopefully wouldn't be too expensive. A true NG AmigaOS might come later, I can certainly agree it wouldn't make financial sense starting with that unless we magically find a huge Amiga investor with deep pockets.

AmigaOS 4 PPC is too expensive to be appealing to all but the most diehard Amiga users but it does come installed on the hardware and there is some software available for it. I suppose RPis could be bought (volume discount?) and resold configured as Amiga Pis for much cheaper than a PPC Amiga. That is the great thing about the RPi price and why it has been so successful for embedded use in other products.

Turrican3 Quote:

I also disagree about appeal of AmigaOS on cheap 68K hardware. Ex-Amiga users left the platform for a reason; they're not coming back for a few additions to a 68K system that was first and foremost about the games. And regarding the latter, I'd argue that AmigaOS is not a strict requirement unless you want to go the WHDLoad route (and even then, you definitely don't need new features)

Many 68k Amiga users left because there wasn't cheap 68k hardware with good performance.

Turrican3 Quote:

I can certainly agree that a Classic Amiga (but not necessarily its OS, which I still love by the way) can look more interesting to the vast majority of the old users. But I don't see this as an issue, since it's far more easier to go the emulation route - perhaps with a dedicated frontend, which is what I'm expecting for the THEA500 - instead of relying on 25+ years old hardware that might fail any day and requires a fairly ancient TV/monitor set. Not to mention it's often not that cheaper to purchase nowadays.

FPGA 68k hardware has modern display device features like HDMI/DVI which displays ECS/AGA/RTG even combining them is some cases. Other modern hardware like USB and memory cards are supported too. The only problem with FPGA is that a choice has to be made between performance (bigger faster FPGA) and price (smaller slower FPGA). The FLEA PPGA Ohm was available for $45 U.S. using a cheap FPGA but without mass production (low volume pricing on components and low volume assembly). More performance and a lower price is possible with mass production and is needed to attract attention, especially from outside the Amiga market.

Turrican3 Quote:

This is why I believe ClassicOS is a dead end and a 4.x port to modern, cheap standard hardware might be a far better solution if (and that's a big if, sure) we want a future for both AmigaOS and its legacy software lineup, i.e. games.

I believe the 68k AmigaOS could be viable on cheaper hardware than Linux can easily support. The target markets would be retro 68k gaming, embedded, hobbyist and educational computing. The Raspberry Pi exposed new markets and gained market share for hobbyist, educational and embedded computing by pushing the envelope of standard low cost mass produced hardware. Would it be better for the Amiga to ride their wave or start a new one? Can the Amiga go lower end yet and with games this time?

A higher end AmigaOS NG with modern features for high performance desktop hardware would be interesting but that market is saturated with mature products which are often free. AmigaOS 4 is a half way transition to that financial dead end. Fortunately, there are some valuable parts which can be ported back to the 68k AmigaOS.

@matthey

Quote:

FPGA 68k hardware has modern display device features like HDMI/DVI which displays ECS/AGA/RTG even combining them is some cases. Other modern hardware like USB and memory cards are supported too. The only problem with FPGA is that a choice has to be made between performance (bigger faster FPGA) and price (smaller slower FPGA). The FLEA PPGA Ohm was available for $45 U.S. using a cheap FPGA but without mass production (low volume pricing on components and low volume assembly). More performance and a lower price is possible with mass production and is needed to attract attention, especially from outside the Amiga market.

FPGA is definitely an interesting route, but I wonder how far can we go from a price/performance perspective.

Since priority in my opinion is to keep price as low as possibile, what kind of performance can we squeeze out of that 45$ FPGA you mention? Honest question, as I'm not that knowledgeable on the subject.

Though we should remember this directly competes with the 2GB RPi4, which costs about 40€ and sports a quad core ARM A72 (to tell the truth I think a RPi 400 would be a better solution, but it costs 3x... still way cheaper than any current available or announced OS4 compatible hardware)

Generally speaking, I think I must be missing something obvious: considering Amiga software emulation works very well, there's a dramatic lack of native modern applications (either productivity or entertainment oriented) while, on the other hand, there's a wide range of cheap, yet decently performing HW like the Raspberry Pi already available that can be targeted, could you please explain what would we gain by keeping the 68K branch alive instead of the more modern and hardware agnostic 4.x / PPC one? I don't get this - again, it's an honest question.

Last edited by Turrican3 on 28-May-2021 at 11:35 AM.

@Turrican3

Quote:

what would we gain by keeping the 68K branch alive instead of the more modern and hardware agnostic 4.x / PPC one? I don't get this - again, it's an honest question.

When even Amikit (totally not anyway connected to A-Eon) says that money is on Classics it should be fairly obvious that large majority what's left of Amiga community doesn't give a flying F about OS4.

@Rose

Quote:

When even Amikit (totally not anyway connected to A-Eon) says that money is on Classics it should be fairly obvious that large majority what's left of Amiga community doesn't give a flying F about OS4.

That's an apples to oranges comparison if you ask me, since the barrier to entry is exceptionally (and arguably unnecessarily) high for OS4.

I for one couldn't care less about OS4... until it was actually available on HW I already owned and for a reasonable price.

Quote:

Turrican3 wrote: @Rose Quote: When even Amikit (totally not anyway connected to A-Eon) says that money is on Classics it should be fairly obvious that large majority what's left of Amiga community doesn't give a flying F about OS4. That's an apples to oranges comparison if you ask me, since the barrier to entry is exceptionally (and arguably unnecessarily) high for OS4. I for one couldn't care less about OS4... until it was actually available on HW I already owned and for a reasonable price.

Remember how people said that there would be tons of new devs if only one could run OS4 on UAE. And I think that 29€ for OS4.1 FE Classic to run on your existing computer is more than reasonable price. Down side is that you will notice REALLY soon that you can do pretty much everything better/easier and not to mention FASTER on host os.

@Rose

I agree, as a person who stopped using OS4 Classic in less time than it took to actually install it (not to mention it was almost a hack) on WinUAE.

That's why I advocate for a *native* RPi OS4 version.

Turrican3 Quote:

FPGA is definitely an interesting route, but I wonder how far can we go from a price/performance perspective. Since priority in my opinion is to keep price as low as possible, what kind of performance can we squeeze out of that 45$ FPGA you mention? Honest question, as I'm not that knowledgeable on the subject.

The fleaFPGA Ohm was $45 for the whole RaspberryPi Zero form factor SBC. The FPGA cost is probably closer to $10 depending on volume discount. This is for a SBC capable of using a 68000+ECS Amiga core which likely could be upgraded to AGA and has 32MiB of memory.



https://fleasystems.com/fleaFPGA_Ohm.html
https://www.indiegogo.com/projects/fleafpga-ohm-fpga-experimenter-board#/
https://www.youtube.com/watch?v=6PU7TAN40Jk

The fleaFPGA Ohm is only scratching the surface of what can be done as far as price. The Raspberry Pi Zero is $5 but lacks ease of use features of the full Pi and Linux doesn't perform well with the reduced spec. The price/performance using an FPGA for the main processor core will never be able to compete here but may be adequate for retro gaming toys. FPGA prices increase substantially as the size and performance grows to accommodate powerful and competitive processor cores. The silicon cost in an ASIC is cheap with mass production though. A single chip Amiga SoC would likely use less than 1$ of silicon. Even a 68060+AGA/RTG wouldn't cost much more and would be faster than when released due to more modern die sizes and higher performance memory. With mass production, we should be able to have very cheap Amiga hardware. The trick to mass production is finding embedded partners who need high volumes of hardware. That's what I was trying to work on for the Apollo core but Gunnar was more interested in playing with his toy by himself.

Turrican3 Quote:

Though we should remember this directly competes with the 2GB RPi4, which costs about 40€ and sports a quad core ARM A72 (to tell the truth I think a RPi 400 would be a better solution, but it costs 3x... still way cheaper than any current available or announced OS4 compatible hardware)

The full Raspberry Pis offer very good price/performance which would be difficult to compete with. Scaling lower while maintaining a good price/performance becomes difficult because Linux doesn't perform well on minimalist systems and low end Pi offerings have targeted low power over performance. Some embedded applications need lower power but others need more performance with the lower price (Power Performance Area trade-offs). I believe the Amiga could turn heads with what it possible on low end hardware but I would rather see it on standard Amiga hardware than foreign hardware where the Amiga is not as efficient or compatible.

Turrican3 Quote:

Generally speaking, I think I must be missing something obvious: considering Amiga software emulation works very well, there's a dramatic lack of native modern applications (either productivity or entertainment oriented) while, on the other hand, there's a wide range of cheap, yet decently performing HW like the Raspberry Pi already available that can be targeted, could you please explain what would we gain by keeping the 68K branch alive instead of the more modern and hardware agnostic 4.x / PPC one? I don't get this - again, it's an honest question.

The Amiga will have trouble getting support for an emulated target, an FPGA processor or hardware or even a wide range of hardware targets. Standard hardware is the reason why the 68k Amiga still has users even as some have moved to emulation and FPGA of that hardware. Mass produced hardware is more attractive than niche market hardware and is relatively cheap for the low end. It is so much easier and cheaper to create Amiga hardware now than when the Amiga was originally designed yet there doesn't seem to be much interest. I guess the current Amiga controllers would rather make a few dollars off the dying Amiga market rather than take some risk reviving the Amiga.

The AmigaOS originally used a flat memory model which uses a relocatable scatter loader to load programs into different physical memory locations. This is simple, fast, saves memory and runs on cheaper hardware without an MMU as sometimes used in low end hardware for the embedded market today. Higher performance hardware uses virtual memory and MMU pages isolate programs from each other. Programs can be loaded at the same fixed virtual address as the pages are changed for each task switch. The advantages are better stability and security due to process isolation and memory protection and virtual memory can be paged to and from drives. AmigaOS 4 adapted the page based SLAB dynamic memory allocator from Linux which has already been replaced by SLUB because of low performance and wasting memory. AmigaOS 4 is caught in the middle. It doesn't have program isolation and barely has memory protection protecting code and maybe the zero page. It has virtual memory paging to drives but most high performance hardware doesn't need it and low end hardware may lack hardware support, reducing performance while wasting memory. The hardware support for paged memory is extensive in high end hardware. Not only is an MMU required but there are commonly multi-level TLB caches, TLB miss hardware acceleration and tags in the caches to reduce the amount of cache flushing on task switches. All of this hardware and a malicious programmer can usually bring a system with paging to a crawl as TLB misses are serviced and/or caches reloaded.

Last edited by matthey on 28-May-2021 at 11:59 PM.
Last edited by matthey on 28-May-2021 at 11:52 PM.

@matthey

What I would like for Christmas is a PPC CPU in a FPGA computer.

_________________
John 3:16

@Trekiej

146 Voters

_________________
John 3:16

@matthey

Quote:

The Amiga will have trouble getting support for an emulated target, an FPGA processor or hardware or even a wide range of hardware targets. Standard hardware is the reason why the 68k Amiga still has users even as some have moved to emulation and FPGA of that hardware. Mass produced hardware is more attractive than niche market hardware and is relatively cheap for the low end. It is so much easier and cheaper to create Amiga hardware now than when the Amiga was originally designed yet there doesn't seem to be much interest. I guess the current Amiga controllers would rather make a few dollars off the dying Amiga market rather than take some risk reviving the Amiga.

Thanks for the explation on the low cost FPGA.

You know, I definitely understand the Amiga is in a very difficult position (and that's quite the understatement), and it's not getting support that easily, if at all.

Plus, I'm not an analyst (and of course I'm not claiming I am), so I have nothing but my gut feeling when I make certain statements.

Having said that, my point is IF the platform holder (which I hope ends up being Cloanto/AmigaCorp, as I have zero trust in Hyperion) wants a future for the Amiga, both as a legacy platform and perhaps as a hobby/lightweight OS too, it needs to target modern, standard, cheap, widely available hardware. To me, this means Raspberry Pi nowadays - or whatever similar mass-market product might end up being successful in the future.

I see no point in keeping support for the 68K legacy OS/code beyond an emulation sandbox, that would be mostly useful for games. Everything else should be moved to ARM as soon as possible, even if it breaks compatibility with old code and/or programs: that's stuff that can run in the sandbox or perhaps it might even get fixed assuming there's the source available (and the willingness + resources to do so)

That's what I would do if I wanted to both keep the retro niche alive (original HW is going to stop working someday) and, most of all, try to secure a viable future for the Amiga.

@Turrican3

Just as on the AmigaOne computers, there's no need to say goodbye to compatibility for lots of old programs, as emulators exist for the Amiga and run well on the Pi, especially Pi4 and Pi400, so if OS4 were to be ported to Pi/ARM in general, most old programs would still work fine.

That would be true if the OS were ported as either BE or LE

Last edited by OldAmigan on 30-May-2021 at 06:12 PM.

_________________
Fred Booth
========================================
A500, A600, A1200 c/w Mediator and 030
AmigaOne and OS4.1
Mac LCII, G4 Powermac running OSX + Amigakit and MorphOS 3.0
Dell Mini 10 Netbook running IcAros and AmigaForever+Amikit+AmigaSys
2006 Macb

@OldAmigan

Survival is the game and until the IP case is solved investment in the platform even from cottage industry case makers like Stephen Jones will not be forthcoming.

Be thankful that things are as vibrant as they are and keep the pie in the sky ideas for the next phase i.e. enticing investment angels who believe in what Mike, Trevor and Matthew want to do with the platform. If that part works then maybe an ISA change is possible if not PPC emulation beckons! Or just stick with the resurgent Classic market.

Last edited by BigD on 30-May-2021 at 08:47 PM.

_________________
"Art challenges technology. Technology inspires the art."
John Lasseter, Co-Founder of Pixar Animation Studios

Trekiej Quote:

What I would like for Christmas is a PPC CPU in a FPGA computer.

Why do you want a low performance high priced PPC CPU in an FPGA? An FPGA can include other logic like the Amiga chipset in an SoC to reduce the cost and save board space but PPC Amigas dropped support for custom chip compatibility anyway. There are PPC FPGA cores available for license likely at a discount considering the demise of PPC. Processor and logic design has been done in FPGAs for decades. Even CBM used small expensive FPGAs for the design of the Amiga custom chips (a $10 FPGA today would have cost tens of thousands of dollars back then).

Turrican3 Quote:

Having said that, my point is IF the platform holder (which I hope ends up being Cloanto/AmigaCorp, as I have zero trust in Hyperion) wants a future for the Amiga, both as a legacy platform and perhaps as a hobby/lightweight OS too, it needs to target modern, standard, cheap, widely available hardware. To me, this means Raspberry Pi nowadays - or whatever similar mass-market product might end up being successful in the future.

What OS has gained double digit market share in the last 30 years without being installed on new/original hardware? Why would any desktop hardware manufacturer want to install a noncompetitive AmigaOS on their desktop hardware, especially if they have to pay for it?

Turrican3 Quote:

I see no point in keeping support for the 68K legacy OS/code beyond an emulation sandbox, that would be mostly useful for games. Everything else should be moved to ARM as soon as possible, even if it breaks compatibility with old code and/or programs: that's stuff that can run in the sandbox or perhaps it might even get fixed assuming there's the source available (and the willingness + resources to do so) That's what I would do if I wanted to both keep the retro niche alive (original HW is going to stop working someday) and, most of all, try to secure a viable future for the Amiga.

Porting AmigaOS to ARM will do little to keep the retro niche alive. Most people who want to emulate the Amiga are already doing so on higher end x86-64 hardware. The cheap price of ARM hardware would partially be offset by the additional cost of AmigaOS due to porting and support expenses.

OldAmigan Quote:

Just as on the AmigaOne computers, there's no need to say goodbye to compatibility for lots of old programs, as emulators exist for the Amiga and run well on the Pi, especially Pi4 and Pi400, so if OS4 were to be ported to Pi/ARM in general, most old programs would still work fine. That would be true if the OS were ported as either BE or LE

An Amiga can be emulated on any OS and on all but the lowest end hardware so why bother porting AmigaOS 4 with noncompetitive desktop features to Pi/ARM at all? Why will users want to use a noncompetitive and expensive AmigaOS on their inexpensive hardware? Why will developers create and support software for a noncompetitive or emulated Amiga? Why would existing 68k Amiga users want to move to AmigaOS 4 on Pi/ARM when they can have better and faster emulation on x86-64 UAE for less money or better compatibility on real or FPGA hardware? How many hardware targets can have good support in a niche Amiga market where updates already take years and drivers are sometimes never completed?