> Unless Intel is struck by lightning and introduces a clean new x86 family
> that disregards all legacy instruction-sets for a new superset.

https://www.intel.com/content/www/us/en/developer/articles/technical/envisioning-future-simplified-architecture.html
https://en.wikipedia.org/wiki/X86-64#X86S
https://www.phoronix.com/news/Linux-6.9-More-X86S

> the Endian-ness question [...] will affect more or less every bit-mask in the
> NDK/SDK, which would mean every application would need to be recompiled.

When porting from PPC to another ISA, recompilation of all binaries would have to happen irrespective of endianness. That's a no-brainer. The real issue would be code that's not programmed in an endian-agnostic way, resulting in a non-working or buggy binary after recompiling to an endianness-switched ABI.

> ARM has instructions that can swap the bits absurdly fast.

I fail to see where bit swapping would be required with a MorphOS compiled for a little-endian ABI.

> SnapDragon is coming with a new series now, made by the same people
> that did the M2 apple silicon.

Gerard Williams (co-)developed Apple CPUs up to M1 and left in 2019 for Nuvia before M2 development started.

@wotanica

So your "best" picks for ARM is something on the same level as a 13 year old Intel CPU and something that hasn't even been announced (and will most likely neither be "open" nor geared at desktop).

Lets say x86 dies in 30 years and stays useable for another 10.... well most of won't care that much.

[ Edited by Kronos 15.03.2024 - 16:55 ]

Quote:

Quote:
Yes, as i outlined it has to. Recompiling and re-doing things like masking is not the same thing.

Quote:

dib graphics is one, blit masking is another, sound processing will likewise be affected.



[ Edited by wotanica 15.03.2024 - 17:04 ]

Quote:

Really, thats what you got out of that?
Me pointing out what is consumer available atm, you managed to think that was the be all?

I forgot for a second why the amiga is in a ditch, but thanks for vividly demonstrating it.

I have created two companies from scratch and have 33 years of software development behind me. Nobody invests 3-5 years of work into a product for a platform, unless there is a long term plan behind the OS they target.

Most people on this forum dont seem to be developers, hence i tried to explain things in a way that would make sense to everyone. Its not a competition about being right, which seem to be a plauge that has infested online dialog the past 20 years. I look for solutions, what can be done. Just listing reasons against is not an argument, nor is knitpicking on "oh its m1 not the m2" when the point is that the same tech will be commercially available.

We can do better than this

[ Edited by wotanica 15.03.2024 - 17:13 ]

Quote:

Well the point is, AMD64 port has started a few years back, so in order to switch lanes to ARM you need to explain why that would be better.
Sofar I only see "it might not be worse" in combination with speculatives "it could get just as good soon" and "it might be better at in few decades".

I don't mind an ARM port, I just don't see any benefits.

>>> the Endian-ness question [...] will affect more or less every bit-mask in the
>>> NDK/SDK, which would mean every application would need to be recompiled.

>> When porting from PPC to another ISA, recompilation of all binaries would
>> have to happen irrespective of endianness. That's a no-brainer.

> Recompiling and re-doing things like masking is not the same thing.

Then you meant to actually write "redone" (i.e. source code adapted) when you wrote "recompiled". But I think that would affect far from "every application".

>> I fail to see where bit swapping would be required with a MorphOS
>> compiled for a little-endian ABI.

> dib graphics is one, blit masking is another, sound processing
> will likewise be affected.

Ah, you mean data endianness. I was thinking about code endianness, that's why I misunderstood.

>> SnapDragon is coming with a new series now [...]. [...]
>> their new Snapdragon will be a general purpose SoC

> something that hasn't even been announced

https://www.qualcomm.com/news/releases/2023/10/qualcomm-unleashes-snapdragon-x-elite--the-ai-super-charged-plat
https://www.qualcomm.com/products/mobile/snapdragon/pcs-and-tablets/snapdragon-x-elite
https://en.wikipedia.org/wiki/List_of_Qualcomm_Snapdragon_systems_on_chips#Snapdragon_X_series

> Its not a competition about being right, which seem to be a plauge
> that has infested online dialog the past 20 years. I look for solutions,
> what can be done. Just listing reasons against is not an argument,
> nor is knitpicking on "oh its m1 not the m2"

No amount of correcting factual errors will be able to stop you from pursuing whatever you think is the solution :-) If it's anything that requires MorphOS source code access, you will have to either turn to the MorphOS team or even join it.

Current MorphOS system has lots of drawbacks/issues:
- Small amount of decent programs
- Limited memory addressing
- No protected execution
- Limited processing power on available HWs
- Outdated libraries
- Lots of time, effort to make the opensource programs to port/maintain (because of endianness problems, incompatible libraries, etc.)
- The ported tools are outdated
- Old Apple computers aren't a good option anymore
- No new HW for MOS available (on reasonable price)

And now, comes the idea to execute MOS on Linux in a virtualized environment? Hm...

Why not shortcut things: Execute those opensource programs (the ported and not yet ported ones) on Linux and all the mentioned problems are solved. And the resources are saved for the applications, instead of unreasonable emulation.

Why not try to make a Linux system to look and behave like a MorphOS? An Ambient like desktop would be a really good thing to see on Linux.
Maybe integrate UAE into the file handler to be able to start old Amiga games, demos, etc. A file handler with .ico files (others are hidden) would be really-really nice to have feature with all the cool programs, which are available on Linux (and no chance to port them).

I think a solution like this could be much more future-proof, since this is 'just' a new shell, a new environment. The MorphOS look and feel can live further without the huge work of porting MOS to a new HW platform. Such system will be no longer just a hobby OS... This can be used even for professional tasks, even for working, with a better user experience.

Quote:
Yes, Linux is already around you can already use it.

Also for the task you suggested (developing a window manager + desktop environment) most probably no access to the MorphOS source code is needed. So it seems any group of people interested enough could start such a project.

Would Linux be much better with a 'MorphOS skin'? I doubt it. All programs despite your DE will continue to use their native toolkit (gtk, qt, ...) anyhow, unless you write special MorphOS versions. Maybe you can also port MUI to Linux, but you'll need to include code + patches to make e.g. Firefox or Epiphany (webkit based browser) use MUI and get a MorphOS-like user experience. From the Linux viewpoint you will just get 'yet another window manager + DE' and from the Amiga viewpoint it would yet another fork.

So more or less you are telling users - just use Linux + UAE. Which of course they already can today. At least for myself there would not be much benefit from a MorphOS-looking Linux over a regular Linux.

Duplicate post....


[ Edited by deka 19.03.2024 - 11:55 ]

Quote:

You're right... Linux will be the same OS. And just another WM... (not Amiga fork). But the 'amigan' way is more fun to use. Fixed place windows, the focusing policy. The huge icons (used to be on older AmigaOS) also could be a very unique feature. GTK or QT should be also personalized to get the requested look.

And yes, it will be never perfect, as the current MOS also isn't. But can be much more prepared to use it every day.

I would not shuffle MUI into this game. Makes no sense and needs lots of work.

[ Edited by deka 19.03.2024 - 12:13 ]

Quote:

One more thought to that:
Such a project is not closely related to MorphOS... but I feel, the evolution of MorphOS is stucked a bit and there is no good direction to go. I think, that would be a new way of keeping things live.

Quote:
While I like the idea of a Linux WM + DE looking and behaving just like MorphOS and it may sound good in theory and logical to do so I think it is quite an undertaking...

You would need to patch GTK-3 and -4 and also QT 5 (probably 6 too) to force the majority of programs to use MorphOS look&feel. Additionally you may need to fiddle around with Xorg and Wayland to make your WM behave the MorphOS way. And you would need to maintain these large out-of-tree patches and adapt them to new GTK/QT versions coming along. This is much more effort than it may first look like.

But true, if this goal would be achieved the end result would be much more prepared to use it every day. Would people pay for using another Linux WM + DE + some closed source software (e.g. Iris). I highly doubt it.

I think once a new multi-core MorphOS capable of utiliting more then 2 GB RAM has been ported to a current hardware platform it will attract new users. If you compare it software-wise to other alternative OSes (like Haiku, RISC OS, ArcaOS, MenuetOS) MorphOS has a pretty good standing and an excellent selection of default apps.

But when this happens we don't know and like to speculate for over a decade now due to it's closed source nature. Personally me too would like to read over some github tracker to check what's going on but we are in Amigaland here and I understand going OpenSource OS-wise would generate too much noise, non-technical dicussion & effort for only little gain.

@ernsteiswuerfel

I accept your arguments about 'Linux DE' vs. 'Separate OS'. MorphOS is a good operating system, even if it has it's issues.

The problem is, that we are in 2024 and everything changes quite quickly in computing. MOS dev. team is too small to keep up with the mainstream OSes. Probably the goal is not that, but... What is an OS for, if you can use it for almost nothing.

Back to my idea: I know, it is not a small task to accomplish. We can reduce it here and there: Maybe it is enough to create a GTK3 theme at first, but to create a new WM for Wayland alone is not a small task. A new file manager is quite similar. And a configurator should be also created.
It isn't profitable at all... Probably nobody would pay for it. :(

I have only one install of a not licensed MOS, but AMITHLON on twenty machines. I really have an issue on the license on MOS, if I buy one I will use and test it on everything I have, without reboot every 30 minutes. I am a retro fan, so I don't care about new hardware...I am a collector of old and very old hardware.

Quote:

You don't need MorphOS developers to do that.

MorphOS needs to keep being MorphOS to stay MorphOS.
What that actually means isn't 100% clear and it can/could evolve into something quite different.
Or it stays on PPC the way it is till it stagnates into oblivion.
Got enough HW to ride that train to my own terminal stagnation

For everything else plenty of YALD to go around as it is.

I think that for future will be much better to unite AmigaOS NG + MorphOS world...
I know, it is some sort of fantasy...

Quote:

I personally just wonder what MorphOS would gain from such deal, everytime I read about this fantasy.

@geit

I agree that from system point of view is MorphOS much much better designed.
But AmigaOS world has more apps, more new games, more modern drivers: besides gfx it is NVME, HD audio.
For another example ( sorry to repeat it again ) our PCI sata.device has only UDMA5 with no S.M.A.R.T, USB has no functional CDDA, ATAPI over USB...

Not to say the AmigaOS community is larger. I have no exact numbers, just comparing life on morph.zone and amigans.net. And MorphOS, even if system developers are great guys, has no person like Trevor... Visions and enthusiasm for public, producing new things...

I like MorphOS, I am using it from beginning of Pegasos 2. Long years it was mine main working / production system. And my last purchase Sam460LE very probably ends with MorphOS like main system. But fact is, that today I spent more time with AmigaOS...

P.S.: and I know, all this stays in category of fantasy. Not due to this practical reasons. But due to political and legal reasons.

[ Edited by sailor 21.03.2024 - 08:00 ]

I‘d say not having a guy with „visions“ is on of the best things about MorphOS.

> I‘d say not having a guy with „visions“ is on of the best things about MorphOS.

You nailed it.

Regarding activity on forums. I think MorphOS is just straight forward and works very stable on all supported plattforms. I know dozens of MorphOS Users who use their machines everyday without participating in any morphos specific online activities. A lot of them use Mac minis or Powerbooks as extension to their classic setup - eg. file transfer, browsing, amiga compatible filesystem creation (usb sticks, hard disks, etc), extraction of huge files, e-uae etc.
But I agree that MorphOS could use some better drivers and more applications and games.