As for the limitations that 68K puts on us, chiefly the limitation of 31-bit (yep) addressing, well that's something that can theoretically at least be dealt with better in a virtual environment with a supporting operating system update. If you rigorously implement the MEMF_PRIVATE concept, processes and tasks can allocate memory in their 31-bit space that's mapped somewhere else in physical memory, meaning 2GB is no longer the practical limit, just the limit per private-memory-aware application. This obviously requires some form of native sublayer to support, but implementing the bare minimum for each physical substrate is a lot easier than trying to create N different architecture versions of the whole thing.

_________________
Doing stupid things for fun...

@Karlos

besides the architectural limitations, what are the use cases for this?

Or simply "why should anyone install it?". Besides you do it because you can of course

@OlafS25

The use case is to have your 68K OS and applications run on whatever hardware you like, without having resorting to a complete guest emulation (e.g. UAE on some other OS) - to be unchained from any particular physical CPU and supporting ecosystem. Want to run it all on a real commodore era system? Fine, no problem. Want to run on FPGA silicon? Fine, no problem. Want to run completely under UAE? Fine, no problem. Want to run on your PPC silicon? Fine no problem. Want to run on your ARM silicon? Fine, no problem. Want to run on your x64 silicon? Fine, no problem. Want a RISC-V base? Implement the bare metal layer (HAL/68K JIT engine) and you are good to go.

I don't have machinations of grandness here, I don't expect AmigaOS to ever be mainstream. What I want is for it to have a means of continuation without the constraints imposed by any particular hardware.

Obviously, there's AROS, but equally I don't especially care about building software for N different architectures. I just want to compile for some optimum subset of the 68K ISA and know it will run on all of the above, as is, provided I respected the OS APIs for everything.

_________________
Doing stupid things for fun...

@Karlos

we misunderstand each other

for me you describe a technical solution, I see f.e. managing private videos and pictures as example for use cases

@OlafS25

I thought you were asking why I would virtualize the Amiga, whereas what you are really asking is what are the use cases for an Amiga, virtual or otherwise, in the first place.

There are not really any sensible use cases. It's all preference and sentiment.

Personally I enjoy working with and on it.

_________________
Doing stupid things for fun...

@Hans

Most users will not mind if the idea is to create Amiga OS4.x for ARM or other hardware that is cheaper true.

But I feel this new hardware is emulated, ans I think that is what the other person was objecting against more so than the fact there are so many PPC haters.

I for one don't mind using my sam flex true, but I also can not justify the price tags in modern world associated with PPC currently with the rise of inflation.

I Digress back on topic :

So if anybody decides to move over the MODERN Amiga OS 4.x? or 5.0 or whatever they call it, then yes, most Amiga user's will be happy with such a concept.

But if it is EMULATED, whats the F'n point?

For emulation I have AmiKit myself, a really awesome software btw IMHO.

But some like Piamiga, ok I can see merit in that I guess.

Others like UAE, FUAE or whatever else exists...

But NONE of these are REAL operating systems as such, and this is what I would like to continue using.

On the retro machine Im using Amiga os3.2.2 (is that what we were upto, forget now)

Again if I want emulation I already have it, we dont need more, as the other user stated, EMULATION MACHINES

SO

Can someone perhaps let me know or notify me when we actually have hardware and operating system solution to the continuing price rises of current hardware.

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

_________________
Courtesy of SAM440Flex & Amiga OS4.1 only
Flex is 800mhz
A1000 with Classic 520 Amiga OS3.2.1
AmiKit 12
MorphOS PowerBook G4 (which can play youtube vids)

https://blitterwolf.blogspot.com

@pixie

Aros could be more user friendly with its installs, regarding the x86 versions. For new users and non geeks, and or tinker types.

AFAIk the Vampire version, as it is dedicated to that hardware, works well, even for users who may not be used to using Amiga and coming back to the platform.

_________________
Courtesy of SAM440Flex & Amiga OS4.1 only
Flex is 800mhz
A1000 with Classic 520 Amiga OS3.2.1
AmiKit 12
MorphOS PowerBook G4 (which can play youtube vids)

https://blitterwolf.blogspot.com

@Gebrochen

ok if it will be fast native up to date with some future
worse retro is not needed

@Karlos

Quote:

I didn't say there was hubris in preferring it. I said the hubris is in thinking you *need* native code, thereby marrying yourself to a given hardware platform. Tying ourselves to particular CPU hardware has worked out so well, wouldn't you say?

I have no idea how many people think we "need" native code. I'd call it a pretty strong "want." My thoughts are that you can use emulation to get us started, but I'd like to be able to run native code. IIRC, Amithlon allowed native code execution, although I don't recall it being used much.

Quote:

Why didn't we just lock down everything else as well? Only one choice of graphics accelerator, one sound, one NIC.

The only reason to not lock it down to specific hardware, is because "planned obsolescence" is too fast. You have a narrow window to use it, before parts become available.

One of the reasons that I'm working on Virtio drivers, is because they open up the possibility of using any graphics card with drivers on a host OS. You could run a barebones host OS with a thin virtual machine (or a full blown emulator), and then use whatever GPU is available. This side-steps the need to write/port any drivers at all (although native drivers are always welcome).

It would require less work than going "Linux driver wrapper" route, like BSD does (BSD has a library that helps port Linux drivers with minimal effort).

Quote:

Firstly, we don't have the resources of Apple when it comes to migrating the platform. We do not have a codebase that is portable enough - by which I mean free of pointer size and endian concerns

We don't need Apple's resources. IIRC, Apple's Rosetta is built on LLVM, which is open-source (and which Apple contributes to). We could piggy back off that, or some other open-source JIT.

Quote:

... Finally, on a JIT platform everything is native anyway. Unless you use a hybrid/hotspot, you aren't interpreting (and they do have their advantages too). Don't think of it as emulation. Think of it as final mile compilation.

If we were going the JIT platform route (which is a worthy idea), then using the 68K architecture for the intermediate byte-code doesn't make much sense. Using a bytecode based on GCC or LLVM's internal intermediate representation would make more sense.

At this point, it's starting to sound a lot like Amiga Anywhere, which ended up going nowhere.

Hans

_________________
Join the Kea Campus - upgrade your skills; support my work; enjoy the Amiga corner.
https://keasigmadelta.com/ - see more of my work

@Gebrochen

as I understand the concept it is not for everyone suited and obviously not for you. Also not for hardcore enthusiasts using amiga all the time for 30 years already with lots of original hardware and only interested in amigaos compiled from original sources like many say. But people coming back and are interested to get new hardware and not buy old hardware for lots of money who is in danger to fail because of age. I think there is a market for it, like with the A600GS concept, how big I cannot judge. On the longer run something like that will be the only option because no old hardware left or unpayable. And the original components not available.

Last edited by OlafS25 on 21-Dec-2024 at 11:33 AM.
Last edited by OlafS25 on 21-Dec-2024 at 10:40 AM.
Last edited by OlafS25 on 21-Dec-2024 at 10:39 AM.
Last edited by OlafS25 on 21-Dec-2024 at 10:36 AM.
Last edited by OlafS25 on 21-Dec-2024 at 10:31 AM.
Last edited by OlafS25 on 21-Dec-2024 at 10:29 AM.

@Hans

Quote:

Hans wrote: , Amithlon allowed native code execution, although I don't recall it being used much.

Full native code was used for updated GFX drivers (and other drivers too)and AFAIK IOSPIRIT product also has native modules.

Almost native code with Martin Blooms byteswapping GCC was used more commonly in opensource ports and other non commercial SW at that time.

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

@Hans

the current amiga community mainly is a retro community. Many even not use something not compiled from original sources, I cannot imagine how they would react on a byte-code concept

@Gebrochen

Quote:

So if anybody decides to move over the MODERN Amiga OS 4.x? or 5.0 or whatever they call it, then yes, most Amiga user's will be happy with such a concept. But if it is EMULATED, whats the F'n point?

For millions of ex Amiga users, there already is no point, beyond playing some old time favourite Amiga games.

Quote:

But NONE of these are REAL operating systems as such, and this is what I would like to continue using.

I've also never had much interest in running AmigaOS in an emulator on Windows/Linux/whatever. Nevertheless, we need to be pragmatic in order to find a way forward despite the very limited resources.

AmigaOS has been devalued over the last few decades. I recently created a special Amiga list for my company newsletter so that Amiga fans can get only Amiga related stuff.** There are currently fewer than 100 people on that list. Okay, i don't own the OS, which reduces how many people are interested in the Amiga stuff I do. Still, it shows you how much the community has shrunk.

Quote:

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....

That would certainly be nice. However, we're likely going to need an emulator to get that across the line in a reasonable time-scale. Either that, or run the OS in a virtual machine, and accept that we start with zero software because we can't run PowerPC or 68K code... The virtual machine is to avoid having to write drivers which is hard even when the manufacturers are willing to give you the hardware docs (which ARM manufacturers tend to treat as top secret).

Quote:

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

True, but AmigaOS 3.2 is a different team, although there is some overlap.

Hans



** If you want to join the Amiga list, then [url=https://keasigmadelta.com/store/subscribe/]sign up here[/url[, and click the link in the welcome email to choose what topics you're interested in.

_________________
Join the Kea Campus - upgrade your skills; support my work; enjoy the Amiga corner.
https://keasigmadelta.com/ - see more of my work

@OlafS25

Quote:

OlafS25 wrote: @Hans the current amiga community mainly is a retro community. Many even not use something not compiled from original sources, I cannot imagine how they would react on a byte-code concept

How did the Amiga community react to Amiga Anywhere?

That said, times have changed and you may find people being more open to it now. If not, try convince Karlos that he's wrong about using JIT compilation.

Hans

_________________
Join the Kea Campus - upgrade your skills; support my work; enjoy the Amiga corner.
https://keasigmadelta.com/ - see more of my work

@Hans

Quote:

Hans wrote: How did the Amiga community react to Amiga Anywhere?

I fail to see how thats really relevant as the "virtual CPU" part was just a bit weird.

The real issues (IMO) were the non existent Amiga relation beyond branding, the low quality of the product, the rather bizarre licensing for 3rd party SW and the limited availability outside running on top of Windows.

And that it was just a minimal viable effort to facilitate an IPO scam that Amino than failed to pull of.

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

@Kronos

besides finding someone implementing something like that, where do you get the software implemented with it? I see there another technology implemented with no real use.

Last edited by OlafS25 on 21-Dec-2024 at 11:36 AM.

@Kronos

Quote:

I fail to see how thats really relevant as the "virtual CPU" part was just a bit weird. The real issues (IMO) were the non existent Amiga relation beyond branding, the low quality of the product, the rather bizarre licensing for 3rd party SW and the limited availability outside running on top of Windows.

Fair point. Amiga Anywhere had zero relationship with the original Amiga beyond the name. That is a big turn off, which dwarfs any debate over compiling to byte-code (whether it's 68K or something else).

When Amiga Anywhere was released, I remember thinking of it as "reinventing Java" and having nothing to do with the Amiga.

Hans

_________________
Join the Kea Campus - upgrade your skills; support my work; enjoy the Amiga corner.
https://keasigmadelta.com/ - see more of my work

@Hans

Quote:

At this point, it's starting to sound a lot like Amiga Anywhere, which ended up going nowhere.

Yes, because it was completely and utterly alien.

68K and 3.x are not.

There are surely better bytecode intermediates you could use or design from scratch, but the entire point of using a suitable 68K subset and execution environment is to allow the software to run on actual 68K silicon and to allow all the OS friendly apps of yesteryear to continue to run without being recompiled. So many sources are lost as it is.

Last edited by Karlos on 21-Dec-2024 at 05:38 PM.
Last edited by Karlos on 21-Dec-2024 at 01:03 PM.
Last edited by Karlos on 21-Dec-2024 at 12:54 PM.

_________________
Doing stupid things for fun...

Quote:

Karlos wrote: Yes, because it was completely and utterly alien. 68K and 3.x are not.

If they had demoed an AmigaAnywhere running old 68k software on an AOS3.x GUI, people would have been happy:



They would even have ignored if it was only software running on Windows.

Quote:

That said, times have changed and you may find people being more open to it now. If not, try convince Karlos that he's wrong about using JIT compilation.Â

I'll die defending this hill. I don't think there's anything daft or outlandish about the idea. You need a low level HAL that embeds the JIT emulation and exposed certain things to the OS for use, e.g. ability to map private memory, devices, etc. Everything else can run within then 68K "box".

On real Amiga hardware the HAL is basically passthrough.

You can't fully modernise without taking away most of the compatibility and things that people genuinely love about the platform.

Last edited by Karlos on 21-Dec-2024 at 01:26 PM.

_________________
Doing stupid things for fun...