Can a RISC-V hardware such as the SOPHON SG2042 host AmigaOS or MorphosOS ?

And if not, why you believe is not an opportunity to have your favorite OS to "updated" hardware ?


SOPHON SG2042
https://milkv.io/pioneer

Last edited by liquidbit on 15-Dec-2023 at 12:02 PM.

@liquidbit

Quote:

liquidbit wrote: Can a RISC-V hardware such as the SOPHON SG2042 host AmigaOS or MorphosOS ?

Theoretically yes, practically no

Quote:

liquidbit wrote: And if not, why you believe is not an opportunity to have your favorite OS to "updated" hardware ?

AmigaOS can't be ported because:

- Legally mess about licenses will make a port impossible.
- Lac of money will make a port unpayable

MorphOS won't be ported because:

- They will port the OS to a hardware with better price to performance ratio and availability


It's funny you didn't ask for AROS, if you would donate such a board to an AROS developer, you might get a Linux hosted version of AROS in less than 6 month.

@liquidbit

Quote:

Can a RISC-V hardware such as the SOPHON SG2042 host AmigaOS or MorphosOS ?

Yep, sure! And I would support such an effort!

OldFart

_________________
More then three levels of indigestion and you're scroomed!

Jeez, can you guys hear yourselves?

There's already a better documented* RISC platform to migrate to, that has excellent 68K emulation, can run big-endian, is affordable, is proven, is maintained and even is popular, so is also servicable/replacable.

The fact that AmigaOS and MOS aren't already ported running on it tells you everything you need to know about why it's not going to run on RISC-V until the devil himself is ice-skating to work.

*RISC-V is very well documented, but I mean the platform as a whole, rather than the ISA.

Last edited by Karlos on 15-Dec-2023 at 08:06 PM.

_________________
Doing stupid things for fun...

@liquidbit

I could see this as viable for MorphOS if there was backing that would be a serious/organized bounty that discussed it in private with the stakeholders at MorphOS first. I have no idea if they would be interested. But, if you had $20,000 or something promised in a bounty prior to any meetings, it would go a long way in discussions.

Bigfoot, recently created drivers for MorphOS for a 10K bounty, so money and seriousness will talk.

MorphOS is really the only viable OS since it has modern applications and tools running natively and has been actively developed for many years. AmigaOS 4 is essentially dead and would rely far to much on old 68k apps and other items making it less appealing and as mentioned no one would touch it because of the sue first ask questions later reputation. Heck I remember when Acill offered Team OS4 either 15k or 25k to port AOS 4 to the Peg II and they basically insulted him (that was on this website in the 90's I think). Then years later they did it for free anyways, lol...

The experience using MorphOS would be a great option for another platform since it the Christmas present of an OS https://amigaworld.net/modules/newbb/viewtopic.php?mode=viewtopic&topic_id=44343&forum=2&start=100&viewmode=flat&order=0#866027

Last edited by Matt3k on 15-Dec-2023 at 08:23 PM.

@Karlos

Quote:

... until the devil himself is ice-skating to work

Just waiting for the "Elfstedentocht" (just 2 more months), but,alas, I'm 'pensionado'.

OldFart

_________________
More then three levels of indigestion and you're scroomed!

@liquidbit

Even though its way more powerful than an X5000 it's still the same price. You just get 63 redundant cores instead of 1 or 3. The point of an architecture switch would be to move away from expensive hardware not run towards it.

Last edited by Rob on 16-Dec-2023 at 11:03 AM.

@OneTimer1

"Legally mess about licenses will make a port impossible"
-Unfortunately I'm old enough to remember the passed conflicts under the Amiga name.

"Lack of money will make a port unpayable"
-I don't agree, there are a lot of money from people which are looking the next opportunity to invest, we saw a lot of waste latelly.

"MorphOS won't be ported because, they will port the OS to a hardware with better price to performance ratio and availability"
..Im in the MOS community since 2003,and as I Peg2 owner I can say that time the hardware running MOS was equivalent with a middle-range PC hardware, now MOS is running on MacMini's which are decades back with today's PC hardware. So they need a new Cell.

"It's funny you didn't ask for AROS"
-I never got the Amiga feeling from AROS, I don't know why, instead I prefer UAE when working on x86 machine.


@OldFart

"Yep, sure! And I would support such an effort!"
-Me too, even in a Alpha stage (financially to the Devs and bying the hardware too for testings).


@Karlos

"There's already a better documented* RISC platform to migrate to"
-You didn't mention the name and a link!


@Matt3k

"But, if you had $20,000 or something promised in a bounty"
-I don't have that money to invest, but 10% of that I could give to a kickstart project.

"MorphOS is really the only viable OS"
-I fully agree.


@Rob

"The point of an architecture switch would be to move away from expensive hardware not run towards it."
-Need to start from somewhere, the price of that RISC-V is not bad regarding its computerization capabilities



My opinion is if Linux can run on this hardware so Amiga like OS could, much easier than Windows anyways..

@liquidbit

You seriously don't know to which platform I am alluding?

Wow.

I am, of course, refering to the Raspberry Pi. You get a whole system and don't have a huge menagerie of third party expansions to cater for. Hell, you can probably even use vendor supplied binary drivers for some of it.

Last edited by Karlos on 16-Dec-2023 at 11:11 AM.
Last edited by Karlos on 16-Dec-2023 at 11:08 AM.

_________________
Doing stupid things for fun...

@liquidbit

For NG systems being ported to anything but PPC you have to explain why RISC-V would be better than the most obvious choice (AMD-64).

Or why it would be better than that other obvious choice (ARM) which itself can only argued to the point of "it might not be worse than AMD".

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

@liquidbit

"But, if you had $20,000 or something promised in a bounty"

I would bet if you asked around you would get more than that pretty easy. Many people would love options with hardware and would put up money, including myself to help it happen.

The MorphOS team is more than just one guy coding it i, I would bet around 10 or more coders that contribute to the OS or appset that will all be impacted by this. So they may have to put it in a timeline, they are very organized with how they handle the OS development, to keep all the coders from falling over each other. So maybe be a 3.20 item or something, again I really don't know the details or how they would feel about it, but I doubt they would drop what they are working on right now to deviate off the roadmap. This is why I would ask around first and get an idea of what you would get, maybe it would cost $100,000 to get it done. I fully believe that 100k would be very doable for this initiative, for bigfoot's driver bounty one person donated 10k by them self, so the money is there but the details will be important. I'm no processor expert either, so I'm not sure if this is the best fit so maybe you keep your options open.

The point is that if you can go to one of the Team with some commitment before you meet, your chances to move on in discussions would be enhanced imho. I would not do a formal kickstarter as I think that would be perceived as too much pressure. Let it organically happen and see what goes from there.

I might also want to see if a MorphOS spreadsheet and accounting program could be added to mix, they are my only 68 apps left that I use, everything else is native MorphOS at this point. If we lose 68k JIT, we need to plan for that as well.

I can tell you that I have been helped over the last 20 or so years by different members of the team, and they are the best bunch of guys I could have hoped to work with. I felt that I was important to them as a user, and many of my issues were fixed over the last 30 or so releases or as part of a beta. That is why I would be as organized and professional with them, so they don't feel I was wasting their time.

@Kronos

Quote:

For NG systems being ported to anything but PPC...

Erm, no. You only need to explain why sticking to PPC today - especially in the context of why it was adopted in the first place - is as moronic as it is ironic.

Except you don't even have to do that, it does an excellent job of illustrating that all by itself.

_________________
Doing stupid things for fun...

@Karlos

Maybe reread the whole sentence until you actually understand it.

The point is that with a port from PPC to anything else (aka the topic of this thread) you would need to find convincing argument why one would choose anything but AMD64, which is near impossible task (finding that argument, not the port).

Sticking to PPC today is bad, going elsewhere is hard and may not be considered worth the effort by those who would have to pull it off. Since neither you nor I are part of that small group.....

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

@Kronos

Moving to x64 would be an embarrassment. The same shade of embarrassment as running on multicore 64-bit PPC, just not as intense because of the better availability/affordability.

Unless, that is, you decide to abandon backwards compatibility with 68k/PPC running outside of a qemu style emulation so that you can address the whole SMP/64-bit issue and, you know, join the current millennium.

But then again, who wants to? If you want a productive x64 experience that's well supported and costs basically nothing, there's a thousand Linux distros. If you want an absurdly fast and efficient x64 experience and don't mind being a fringe user of a system almost nobody has heard of (and being a PPC/NG user, let's face it, that's you by the bucketload), there's MenuetOS. Others to.

In short, moving to x64 with the intent of breaking compatibility with legacy applications so that you can focus on being a better x64 OS, throws away most of what makes MorphOS what it is, and leaves you as a tiny fish in a vast ocean of more mature and frankly better operating systems for the same hardware. One of more of which you'll just end up multibooting with in order to do anything worth doing that isn't twiddling icons, wbprefs and screenshotting the result to post on forums.

And while I personally wouldn't give a crap (my PC is almlst retro now), that I think is the real reason the x86/64 MorphOS is still just a nerd dick tease.

Moving to ARM doesn't quite have the same issue. Even if you are rocking a single core in a 32-bit big endian memory model like Emu68, who cares? Underutilizing the hardware here isn't quite as humiliating as watching a gaming rig tonk along with a basic VESA frame buffer on a single core of the CPU.

ARM also satisfies the need to be "intel outside" for your average zealot too.

_________________
Doing stupid things for fun...

@Karlos

ARM would be the same "bad" with just some extra sprinkles on top of it.

Whether I underutilize a single core of architecture a or b while running endian swapping 31bit code makes no differences.

If I cared about "Intel outside" (which I never did) going with AMD is kinda the cleaner way of doing it.

Plenty of low power AMD CPUs around and pretty easy to combine them with the best supported GPU or replace them with something stronger if a "QBox" ever becomes a reality.

So yep "it might not be worse than AMD" is the only argument you made sofar.

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

@Kronos

I think it is. After all, running MorphOS fluidly on old Mac hardware that OSX struggled on is something cool and has nerd bragging rights for sure. MorphOS flies like crap off a shovel on an old G4 MacBook, making apple look like crap on their own turf. I loved that. What's that strapline? "Experience the natural beauty of resource efficiency" ? That was it.

It's rather hard to make that claim when running like crippled gimp on contemporary x64 equipment. Suddenly it's in a completely different league and is outclassed at almost every turn. The OS would have to evolve in a way that needs to solve all the problems your average blue troll spent years mocking thr average red troll over.

While this may be technically true on ARM, you don't have the whole PC master race to put up with and most people using things like Pi are already totally underutilizing it and wouldn't know what the damn thing was capable of anyway.

You get a nice standard hardware package that's easy to support and extract as much as you can out of. It's affordable, nerdy, low power, can be retrofit nicely into an old Amiga wedge of you wanted. It's cool in a dozen ways x64 just isn't. And don't get me wrong, I like both.

I actually don't care which it moves to. I happen to think ARM would be a less bitter pill for the userbase. But there's one thing for sure. Sticking to PPC is not a viable proposition if you want to have a future of any kind, because PPC doesn't.

_________________
Doing stupid things for fun...

@Karlos

Quote:

Karlos wrote: you don't have the whole PC master race to put up with

HINT:
None of them would give a flying f### anyways.

The "it runs Windows and we can't compete with that" was downright retarded back in the Amithlon days, and it surely didn't get any smarter in the 22 years since.


It is really what do we (or better those who might do a non PPC MorphOS) want to achieve?

Given their past stances I'd say they would prefer hitting the bare metal of an old Radeon over glueing themselves to some binary blob of unknown quality and longevity. I suspect they also care about more about getting the most performance out of that single core then about how many other cores are "wasted".

It is also no secret that a proof of concept AMD port does exist so that seems the most likely and sensible way to follow.

If the turn around and to go ARM (or RISC-V for that matter) I will get myself some of that HW and stay for the ride as long and wherever it takes.

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

@liquidbit

funnily I boot Aros with WinUAE

@Kronos

I know the x64 port exists, and has done for a long, long time. Which is why I can only assume the reasons for not migrating by now are rooted in the obvious: it's gone from the best OS for a bunch of obsolete old hardware nobody gives a crap about anymore to a truly mediocre one unless it goes full SMP/64 bit at the same time. Which requires solving some hard problems and likely breaking backwards compatibility which has always been one of the key strengths of the platform for former Amiga users. Which are the only users, unless there are some old Apple PPC converts. Lol, no, couldn't type that with a straight face.

_________________
Doing stupid things for fun...

@Karlos

One can lead a horse to water...

_________________
All the way, with 68k