@NutsAboutAmiga

You lost me, I have no idea what you’re talking about… what I suggested a few posts ago was an adaptation of OS3 specifically for Emu68 Pi “standalone”, ie just a Raspberry Pi, without amiga hardware attached, Petunia has no part to play in that.

_________________
B5D6A1D019D5D45BCC56F4782AC220D8B3E2A6CC

I'll take a standalone new RPi solution running 3.x on Emu68k than any new PPC hardware any day.

Look at the 68K lightwave benchmarks and compare the cost/performance ratio of the PiStorm/CM4 against PPC hardware costing many multiples the price. It's embarrassing.

The sum total of PPC software for OS4 versus the sum total of 68K software is also not a winning ratio. It's time to end the ridiculous charade. If OS4 is going to survive it needs to exit PPC. Literally almost anything else makes more sense now.

OS4 on ARM, implement an efficient 68K JIT. Done.
OS4 on 68K. Run under UAE or Emu68 or on 68020+/RTG. Done.
OS4 on 68K ubder JIT on x64 (Amithlon style). Done.

But what about all my PPC software? (You know that vast corpus of can't live without it stuff ) Well, seriously. How much of it is actually dependent on PPC? Almost none. Including OS4 itself.

PPC is the deadest of all dead horses.

_________________
Doing stupid things for fun...

@Karlos

What would be the point of bogging down Emu68 with OS4?

_________________
B5D6A1D019D5D45BCC56F4782AC220D8B3E2A6CC

@kolla

I'm not a hater.

OS4 isn't the problem, really. It never was. It was fine in everyday use for me, decent backwards compatibility and decent 68K performance. I don't have an issue with it at all (this shouldn't be a surprise as I was a contributor). The problem pretty much has always been the hardware. The BPPC/BVision was ok provided you stuck to RTG modes and didn't use compositing, but those are of course features you'd want to use. The A1 was faster and, when it worked, a nice experience. But we all know the problems it had, and one day with no warning at all, mine joined them in silicon heaven.

Since then we've just had more and more hardware that has issues or can't be adequately supported or goes significantly underutilised and just costs more and more. If you want a decent OS4 experience today you have to source hardware as rare and expensive as unicorn turd and expect to only be able to use a fraction of it. It's just absurd.

I have no doubt that OS4 recompiled for 68K would run extremely well on PiStorm.

Last edited by Karlos on 21-Oct-2023 at 07:14 PM.

_________________
Doing stupid things for fun...

@Karlos

Quote:

Look at the 68K lightwave benchmarks and compare the cost/performance ratio of the PiStorm/CM4 against PPC hardware costing many multiples the price. It's embarrassing.

Still slower than nearly cheapless MacMini G4 or PowerMac G5 with MorphOS.

@Karlos

Quote:

I have no doubt that OS4 recompiled for 68K would run extremely well on PiStorm.

Changing and compiling it to another dead instruction set,
just so it can be emulated on PiStorm?? Really…



Wont it be better emulate PowerPC instruction set, at least then you wont need to recompile everything, native target should be ARM instruction set.

Last edited by NutsAboutAmiga on 21-Oct-2023 at 09:44 PM.
Last edited by NutsAboutAmiga on 21-Oct-2023 at 09:00 PM.

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

@rzookol

Which has nothing whatsoever to do with the cost of entry for OS4 on new hardware.

Look, MorphOS is a great alternative but why it sticks to decaying fruit cores when it was demonstrated booting on x64 I can't even remember how many years ago, is beyond me.

For years, it's amused me whenever the merits of MorphOS are touted over OS4 when it continues to share the single biggest flaw with OS4: being tied to PPC. Absolutely feck all of any merit in either ecosystem is incapable of being ported to another architecture because it's almost all written in C or higher level languages.

So sure, it's a more cost effective and polished PPC experience for an Amiga user, no doubt, but in the end a more polished turd is still a turd.

_________________
Doing stupid things for fun...

@NutsAboutAmiga

Expand your thinking a moment. It's not about compiling for 68K, it's about using 68K as a binary Lingua Franca. How it is executed is almost irrelevant.

Having a single binary format to support while allowing hosted execution on different hardware architectures would be a boon for the platform. It has far more potential than sticking to the 32-bit subset of a single core of a multi core 64-bit PPC you spent a ton of cash on. Now that truly is moronic, you could liken it to driving around in an expensive V8 that you've hacked to run on just one cylinder using a two stroke cycle.

Quote:

Wont it be better emulate PowerPC instruction set

No. This is the exact thing we need to just put behind us and deny all knowledge of in the future, if there even is one ...

Emulating PPC well is hard. The end result isn't going to be stellar. For the same host CPU effort, you can have much better behaved and I dare say performant 68K emulation.

If you have the HLL sources for something then recompiling it is a far better idea. Whether that's to ARM or to 68K. Literally anything but the bloated deadend that is PPC.

Last edited by Karlos on 21-Oct-2023 at 10:50 PM.

_________________
Doing stupid things for fun...

If OS4 and MOS stick to PPC they are going to die, slowly, in the dark. They'll never be "retro" or "classic", simply because they never had a big enough user base to begin with. The 68K scene is way more active.

_________________
Doing stupid things for fun...

@Karlos

Quote:

Karlos wrote: If OS4 and MOS stick to PPC they are going to die, slowly, in the dark. They'll never be "retro" or "classic", simply because they never had a big enough user base to begin with. The 68K scene is way more active.

Neither is any fast "68k", whether it's on a rPI, Vampire or UAE.

"68k is more active", sure but thats for the most part games targeting an 1MB A500 since thats about the best a team of 1-3 hobbyist can do without the result being "meh" either by being a port or just underwhelming.

All fast "68k" can do is run obsolete SW faster and easier making them a faster turd which is still a turd just with the potential of a real big mess.


-> anything Amiga that isn't related to playing A500 level games will eventually die, just like anything that doesn't run on an C64+floppy is pretty much dead in that scene.

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

Quote:

Karlos wrote: Since then we've just had more and more hardware that has issues or can't be adequately supported or goes significantly underutilised and just costs more and more. If you want a decent OS4 experience today you have to source hardware as rare and expensive as unicorn turd and expect to only be able to use a fraction of it. It's just absurd. I have no doubt that OS4 recompiled for 68K would run extremely well on PiStorm.

Cheers for giving me a chuckle this morning, sadly that sure sums things up from what Ive seen just watching from the sidelines all these years (you've got to laugh or cry)!

Having OS4 running on actual Classic Amigas via cheap PiStorm (or other 68K options) would make it much more appealing to a lot of folks including me. Not everyone has tons of desk space for a myriad of Amiga like systems …. or has the time to keep them all up-to-date & in good working order.

_________________
A1200+Mediator+Voodoo3+040+130mbRAM+0S3.9
A2000+Supra28mhz+9mbRAM+OS3.2.2, CD32 & WinUAE

@Kronos

Quote:

All fast "68k" can do is run obsolete SW faster and easier

Absolute nonsense and your lack of imagination with this statement is a great indicator of the problem. What you have said applies to a real physical 68K processor from the motorola era. You are thinking of the imitations of 68K as a hardware platform and not thinking about it as a basic binary format for execution on varying platforms.

In that sense the only real limitation it has is already shared with existing "NG" PPC, which is the propensity to stuck with 32-bit addressing. In that respect they are all obsolete.

You will almost certainly not make a fully backwards compatible PPC NG OS that scales to 64-bit SMP seamlessly in the other direction. You will have to choose one path or the other.

Take away seamless backwards compatibility with "obsolete SW" that outnumbers your NG software catalogue 50:1 (conservative), and what have you got? A quirky little OS that runs almost nothing on a dead end platform that you could run Linux PPC (there are at least 10 distros still supporting it) that has far more software available and makes far better use of the same hardware. You wonder why it hasn't moved to x64 yet? All that obsolete software couldn't be a reason, surely ...

In some respects, the A1 and Peg 2 where the best new hardware for either OS since they had a chance of being able to fully utilise it. MorphOS on a multicore G5 is just as silly as OS4 on pretty much every multicore CPU based golden boat anchor that's come along since. You can argue about degrees of maturity and backwards compatibility and source heritage, but in the final analysis, the only advantage MOS has over OS4 is that you can, with luck, find a machine that someone else (arguably with more sense) threw out in favour of x64 because their platform had the sense to migrate, then eke a few more years life out of it. Which in turn doesn't require the sale of your first born to fund.

_________________
Doing stupid things for fun...

@ppcamiga1

Would you STFU no one actually cares what you think.

_________________
Support AROS sponsor a developer.

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

@kolla

Quote:

kolla wrote: What if OS3 team suddenly decide to target Emu68 “standalone” as supported hardware?

There’s an OS3 team now?

_________________
All the way, with 68k

@kolla

Quote:

kolla wrote: @Karlos What would be the point of bogging down Emu68 with OS4?

Many reasons:
1. It’s much harder to bog down a virtual CPU because the underlying hardware can be cost-effectively bumped up
2. AmigaOS 4 is more portable than Amiga OS 3
3. Hyperion are back-porting OS 4 bits into their OS 3 cash-grab
4. AmigaOS 4 has a modern-ish web browser
5. Eliminating the need to work on two fairly similar operating systems
6. Transparently run 68k software from inside OS 4
7. They could even not give up on their latest (relative terms) feature and continue the work on multi-core support using multiple emu68 processes
8. It’ll piss off @ppcamiga1 - This one has to be my favorite. I’d even go against my own principles and buy AmigaOS 4 for emu68 just to briefly bask in the sweet sweet schadenfreude.

_________________
All the way, with 68k

@kolla

Quote:

kolla wrote: @Karlos What would be the point of bogging down Emu68 with OS4?

Many reasons:
1. It’s much harder to bog down a virtual CPU because the underlying hardware can be cost-effectively bumped up
2. AmigaOS 4 is more portable than Amiga OS 3
3. Hyperion are back-porting OS 4 bits into their OS 3 cash-grab
4. AmigaOS 4 has a modern-ish web browser
5. Eliminating the need to work on two fairly similar operating systems
6. Transparently run 68k software from inside OS 4
7. They could even not give up on their latest (relative terms) feature and continue the work on multi-core support using multiple emu68 processes
8. It’ll piss off @ppcamiga1 - This one has to be my favorite. I’d even go against my own principles and buy AmigaOS 4 for emu68 just to briefly bask in the sweet sweet schadenfreude.

_________________
All the way, with 68k

@purists

My A4000 has a ZZ9000 for RTG + Ethernet. The hardware solution uses 2 x ARM Cortex A9 cores. Does that make my A4000 unclean?
Who is left out there that can bless and make kosher the non-PPC hardware for the Amiga platforms. Does Sensei Dave Haynie wield such power?

Also while I got you: If I’ve written things in PHP, Python, Node.js, or Wasm on Linux (x86-64), how much time would it take me to learn the PPC versions of PHP, Python, NodeJS, or Wasm?

_________________
All the way, with 68k

@agami

Quote:

migaOS 4 is more portable than Amiga OS 3 3. Hyperion are back-porting OS 4 bits into their OS 3 cash-grab

This is it. People seem to be living under some insane delusion that because OS4 was difficult to incept from 3.x, that this is somehow reciprocal. The effort was not in moving from 68K to PPC, it was moving from assembler to C. Now that that's done, you can port to any other 32-bit big endian target relatively easily. You can even port to any other target, albeit with more challenges and loss of compatibility with respect to data structure layout.

_________________
Doing stupid things for fun...

@agami

Quote:

agami wrote: @kolla Quote: kolla wrote: @Karlos What would be the point of bogging down Emu68 with OS4? Many reasons: 1. It’s much harder to bog down a virtual CPU because the underlying hardware can be cost-effectively bumped up

Can it? Emu68 works on Pi5 already?

Quote:

2. AmigaOS 4 is more portable than Amiga OS 3

Really? How do you know? Do you have any prior examples of this?
And what does it matter if the target is 68k?

Quote:

3. Hyperion are back-porting OS 4 bits into their OS 3 cash-grab

Yes, bits and bobs, and it's a bumpy road! Mostly it has been reimplementing OS4 features on their OS3 counterparts, though a few pieces have been sort of backported, like C:List.

Heck, the current OS team struggles to make even the Prefs programs using the much superior ReAction that was supposed to make all such tasks much easier! Apparently, it isn't, and instead gadtools Prefs have been updated. Thankfully.

Quote:

4. AmigaOS 4 has a modern-ish web browser

Heh, ha ha. Even the IP stack on OS4 is way behind that of OS3.

Quote:

5. Eliminating the need to work on two fairly similar operating systems

They would still have to do that, OS4 would not be usable on real 680x0 which remain targets for OS3.

Quote:

6. Transparently run 68k software from inside OS 4

What relevance does that have when the hardware _IS_ 68k?

Quote:

7. They could even not give up on their latest (relative terms) feature and continue the work on multi-core support using multiple emu68 processes

What "processes"?

Quote:

8. It’ll piss off @ppcamiga1 - This one has to be my favorite. I’d even go against my own principles and buy AmigaOS 4 for emu68 just to briefly bask in the sweet sweet schadenfreude.

So far the only valid reason.

But this is all unicorns and rainbow poop - OS4 is not coming to 68k, and from what has been presented at latest shows, even 3.3 is years away.

_________________
B5D6A1D019D5D45BCC56F4782AC220D8B3E2A6CC

@Karlos

Quote:

Now that that's done, you can port to any other 32-bit big endian target relatively easily.

Then Emu68 is a moot point, as they should rather target native RaspberryPi.

Last edited by kolla on 22-Oct-2023 at 04:04 PM.

_________________
B5D6A1D019D5D45BCC56F4782AC220D8B3E2A6CC