When will a new version be released?

As for the new version of the system, I would prefer it to support new hardware. For example x86, AMD64.

Honestly?
I really hope NOT an X86 AMD64 version
Some versions have been shown but after so many years if they had wanted they would have already done it
Supporting X86 hardware is impossible too many variations between video cards motherboards processors
Impossible to support all this hardware
we would be forced to look for the right SPECIFIC card anyway
at least those are the Macs

@cip060

- GFX cards would be the same ones already supported in PCIe(and some AGP ones) Macs and X5000/SAM460, so no real problem. Actually it would be slightly better as there would be no issue of needing a 2nd card to access FirmWare/BIOS/BootMenu.

- I don't think it would need to be super specific to a certain board. AFAIR Bigfoot used a MSI B450 board and I'm almost certain that the variations to other B450 would be less than between different revisions of some Macs all sharing the same model number.

- B450 boards with AM4 CPU can still be bought new and GFX cards are easy enough to find

Even if you would need be limited to a few specific boards and a Ryzen3600 (which Bigfoot used) that is still easier to find than a working PCIe G5 which have become slim pickings over the past few years.

So no, HW is not the issue, what would be an issue (IMO) is that the underusage of the HW would be even more apparent when everyone has AT LEAST 4 cores and 8GB of RAM.

Hello everyone, first intervention here even if I have been following this site and this forum for years. Forgive my English, it is not my native language and I mainly use Google Translate.

In short, all these years, I used MorphOS on Pegasos I/G3, and now on my PowerBook 15" 1.67/DDR 2. I notice something really annoying: we arrive at a really mature software solution (thanks Jaca and its great browser that opens up all modern uses), but I feel completely limited by the hardware. Just yesterday, I was using my PowerBook to surf and the slowness was driving me to despair. Which ultimately leads me to only start my PowerBook occasionally, while with it I could meet 95% of my needs (especially via the internet and we have the browser for that).
As for finding a new machine, I am frustrated: no new more powerful laptop, the iMac G5 iSight are beautiful but hard to find, the PowerMac G5 are just as hard to find but bulky and not very energy efficient... And in the end, for what real gain?

It is really time to change platforms. The platform x86 has the advantage of being very widespread and sustainable, but finding the right configuration, especially on a laptop, seems difficult to me and potentially a few hundred Euros to pay. In my humble opinion, the Amiga spirit is above all hardware and software optimization. I see a solution that corresponds to that: the Raspberry Pi! If we detail:
-the Raspberry Pi is roughly equivalent to the G5 since the Pi 3 B+. With a Pi 4 or Pi 5, we gain in performance.
-It's small, cheap, easy to find, ideal for an OS that is considered by many as a hobby.
-With the Pi 400, we have a good reminder of the past like the Amiga 600.
-No need to bother with multiple configurations, it facilitates the development of drivers.
-It's energy efficient.
-It's well documented.
-It would allow to expand the user base easily.

Add to that the fact that the ARM platform is sufficiently durable, that we could have a widely available ARM laptop with MacBooks or Qualcomm machines (not to mention tablets with keyboards), that we have high-performance platforms with the Mac Mini Silicon...

I sincerely hope that the MorphOS Team is studying this path and that the development time since the demonstration of MorphOS X86 corresponds to a change of architecture towards ARM. Because personally, I no longer have the means in recent years to spend €700 or more on a hobby that does not benefit my family, while €100 on a Raspberry (or even nothing, I already have a Pi 4) I would do it without a problem. As for the MorphOS license, it has been at a really reasonable price for years.

In short, it was "my 2 cents".

@Pistouillette

All ARM boards have on chip GFX that would not only need new drivers to be written, but are also often pretty closed
and only supported via binary blobs under Linux.
Once you go into more obscure solution allowing discrete GPUs you end up with much higher price tags.

Depending on the exact supported HW you could get a a barebones (board,CPU,RAM) for 150€ and a fraction of that for used HW.

Super cheap AMD laptops do exist, but support would of course be an issue (just like with any ARM laptop).

Getting that old AMD port into useable state seems feasible, while an ARM port would be a start from scratch with serious roadblocks to get to the same state.

Quote:
But is it enough still? If I get a new MorphOS machine, I would like it to last for a long time in performance wise too. I have Raspberry Pi 400 myself, but with Linux it feels quite sluggish already, which doesn't inspire me to use it at all even though I like its looks.

Quote:
I had hard time to find RPi 400 couple years ago, they were out of stock everywhere and had to order it from another country. Maybe the situation has recovered now?

Quote:
That's the only thing I'm happy with it ;) White and red case and mouse remind me about Amiga.

Quote:
Well, they're updating the hw time to time.. don't know how backwards compatible they are from drivers side... and there might be those nasty binary blobs that are too closed for non-mainstream operating systems...

It seems to me that the Raspberry Foundation chose the Broadcom SOC of the Pi because it used the Videocore VI (Pi 4/400) or Videocore VII (Pi 5) which is documented and for which there is an Open Source driver. There has even been work on Vulkan on this subject. Wouldn't this allow, based on the existing Open Source code, to build a driver for MorphOS? If necessary, with a bounty to finance the work. This has been possible recently for older AMD GPUs. And in any case, for new x86 machines, it will also be necessary to develop the drivers (I am thinking of laptops and integrated AMD GPUs). For the other drivers (bus, audio, USB, network, wifi, etc...), the workload would indeed be enormous. But is it impossible? Here again, the work seems to have been feasible for Macs, X5000, partially for the x86...

In terms of performance, the Pi 4 is significantly faster than a G5 and for a Pi 5 the difference will be even greater. We would really gain from the change. Especially since MorphOS is much lighter than many Linux distributions. For everyday use (office, internet, drawing, light photo editing, a few not too demanding games), it would be enough for the majority of MorphOS users, right?

As for fears about availability, there was indeed a difficult period, but now it is easy to find. Easier than an old PowerBook, iMac iSight or PowerMac G5 in good condition and with the right graphics card. And it is also very common second-hand. Not to mention the reliability on Apple PowerPC hardware that is approaching 20 years old.

For x86 hardware, my fear is that we are on very specific configurations, which will soon encounter the same problem as Macs: not found new soon.

Quote:

Well my last info on that is it still goes down to a binary blob, but even if it was 100% opensource it would still a different architecture compared to the Radeons so surely some extra work.

Also since we don't have PCI(e) slot everything would either be supported onboard or via USB dongles. All to get inferior single core performance (thats all that matters for now) even when compared to dumpster dived AMD CPU from 2010 (not suggesting that those would or should be supported).

All these ARM vs AMD discussions tend to boil down to "it might not be worse" with some sprinkling of "I think it would be kinda cool, just can't tell you why".

Now if we think far far ahead to ARM or AMD laptops you run into the same problems for general IO, power management and touchpads with the slight difference that integrated GPUs on ARM will not be the same as on the rPI and most likely not fully documented while on AMD you might get to the point where the lowest iGPU shares it's architecture with some discrete Radeon already supported.

> It seems to me that the Raspberry Foundation chose the Broadcom SOC
> of the Pi because it used the Videocore VI (Pi 4/400) or Videocore VII
> (Pi 5) which is documented and for which there is an Open Source driver.

The original decision to use the BCM2835 was more despite of the VideoCore IV than because of it. Broadcom delivered the SoCs at cost, and Raspberry Pi Foundation's Eben Upton had previously worked at Broadcom's VideoCore development team. The VideoCore GPU drivers only became fully open source a while after the release of the first Raspberry Pi.

https://www.theregister.com/2012/10/24/raspberry_pi_broadcom_soc_drivers_now_fully_open_source/

You all miss the point.

Platform swap is not just about the cheap target hardware. It is about 64 bit support and changing the entire System API to support multi core/cpu and memory protection.

It makes no sense to just swap the hardware. This will still leave us in the 1 CPU with max 1,8 GB RAM hole with all its limitations.

Since every step as a result means loosing software compatibility this needs to be done in one go.

Quote:

Does it really?

AFAIR bigfoot's proof of concept did run in 31bit big endian with at least the possibility of a seamless PPC EMU.

Anything little endian, 64bit and multicore could and should be build out as a "QBox" directly in and around the Quark kernel.

I fear "done in one go" is a non starter with the limited resources.

> on AMD you might get to the point where the lowest iGPU shares
> it's architecture with some discrete Radeon already supported.

Currently with newest supported GPU architecture being TeraScale 2, this would be AMD SoCs from 2011:

https://en.wikipedia.org/wiki/List_of_AMD_processors_with_3D_graphics#Lynx:_%22Llano%22_(2011)
https://en.wikipedia.org/wiki/List_of_AMD_processors_with_3D_graphics#Sabine:_%22Llano%22_(2011)
https://en.wikipedia.org/wiki/List_of_AMD_processors_with_3D_graphics#Brazos:_%22Desna%22,_%22Ontario%22,_%22Zacate%22_(2011)
https://en.wikipedia.org/wiki/List_of_AMD_processors_with_3D_graphics#Brazos:_%22Ontario%22_and_%22Zacate%22_(2011)

Few advantages I see going for ARM on Raspberry:

- machines released ten years ago are still produced/supported (the PI B+ from 2014 is still being produced)
- it's really cheap (no hardware barrier), plus already widespread (61 millions of PIs have been sold since launch)
- I guess it's the most opened ARM platform (gfx drivers are open source,...)
- the Raspberry company is opened to more OS on its platform (there is already a version of RiscOS for the PI), and Eben is fan and has fond memories of the Amiga
- It's fast enough already
- I think there is room for a light/fast multimedia OS like MorphOS on these boards
- Imagine what you could do with these GPIO pins :)

I guess it's the platform with the easiest (like free) entry fee for MorphOS. It doesn't mean it would magically work though.

@Leo

Minor advantages padded with quite a lot of "I think it would be kinda cool, just can't tell you why".

On the other hand we would be talking about scarifying both some peak performance and whatever work bigfoot has already done.

Don't get me wrong, if they go ARM I will follow. I just don't see any compelling reason for going in that direction.

Personnally, i have 2 laptops with Ryzen7 and Radeon waiting MorphOS (i hope MorphOS could start on them one day maybe...)

Quote:

Yes, I'm sure a Hollywood GPIO plugin is exactly what all MorphOS users have been waiting for.

Quote:
As a MorphOS user I'd love to play with some GPIO. I never used nor felt the need to use Hollywood though. But I guess you know better the typical MorphOS user profile than I do.

[ Edited by Leo 16.09.2024 - 14:47 ]

@Leo

Let me put some sprinkles on that

/s
/s /s /s


*yummy*


Not that you couldn't get some IO pins on a PCI(e) card or USB dongle and then write some drivers for it.

Quote:
Exactly.
1st step port and release MorphOS 32bit for x64 systems.
2nd step make some MorphOS 64 with multicore and what else is missing and put MorphOS 32 bit into a box/virtualize it.
The boxing idea is there since 25 years now. Time to use this idea eventually (not only the hardware is ageing, we all do!)
One in a go is nice, but probably unrealistic.

But hardware is aging and cpu grunt often not sufficient (albeit Wayfarer became impressively fast!). MorphOS itself and the applications are better than ever though (I actually can do much of my daily work on MorphOS these days!).
Pity the hardware is holding us back.

well, the traditional desktop market is shrinking and intel has lost its edge. I dont think x86 will survive much longer as ARM has set its foot into the classic PC domain with Mac shifting away from x86.

I'd rather prefer a cheap and reliable platform that'll be still supported in 10 years. From the user's perspective a raspberry sounds really nice as it is cheap and has a huge userbase (with a lot of techies and nerds just like the amiga in its heyday who love to fiddle around with their machine).

Releasing MorphOS on a x86 machine wouldnt generate as much interest and buzz as a release on the Raspberry Pi. MorphOS is way better than RiscOS and more leighweight than linux. It's a win win situation imo.

You also have to bear the question, if that specific x86 Mainboard will still be available in 10 or 15 years after the release. Today we can still buy PowerMac G5s and Powerbooks as those systems are all defined by a standard configuration and thus easily identifiable. On x86 you will have to look for a specific mainboard which mets the exact requirements for morphos. I dont think this is going to be an easy task. Resourcing compatible hardware will be way harder. Alternatively the MorphOS team will have to keep adding support for more x86 hardware on a regular basis.

[ Edited by Cego 18.09.2024 - 11:09 ]