I know that the MorphOS Dev. Team members have decided that porting MorphOS to x64 is the direction they want to go for the future next step of development, but I would like to begin a discussion to gauge the interest, and potential for support, to port MorphOS x64 to the ARM architecture, once the Dev. Team has finished the first release of MorphOS x64.

My opinion is that MorphOS will probably always remain only a hobby OS, and never really compete 1:1 against Windows, MacOSX, or even Linux. I would love to be proved wrong on this point, and see the popularity of MorphOS grow into something that could compete with Linux distro's, but that is only a "pipe dream", IMO.

If MorphOS will always only remain a hobby OS, I believe it could be most successful, on cheap hobby ARM devices, unless the support for MorphOS on x64 hardware can easily be expanded to hundreds of consumer desktop and laptop/netbook/tablet computers. With the more likely support of x64 hardware being limited to one or eventually 2 or 3 different CPU/GPU & chipset combinations, due to limited resources to write device drivers, I'm guessing that most average computer users will not have the correct combination to run MorphOS x64.

If we could get MorphOS x64 ported to the Raspberry Pi family of boards, which appear to all use compatible components with each other, we would have the possibility of adding hundreds of thousands of potential new MorphOS software developers and tinkerers, since there are already millions and millions of Raspberry Pi boards sold, and many million more likely to be sold between now and when MorphOS x64 is finished, and porting of it to the ARM architecture could begin.

So, finally to my question; how many members here would be willing to pay into a bounty to port MorphOS x64 to the Raspberry Pi 3 (or 4, 5, 6, by the time a port could probably begin), or just the ARM architecture in general? Unless such a port could be done by 3rd party ARM developers now, and they could work side by side with the existing MorphOS Dev. Team members to do a MorphOS for ARM port while the x64 version was being developed at the same time, I would not want to (nor do I think the team would agree to trying to do it themselves) take away the focus of the existing MorphOS Dev. Team members, from their work to create the x64 version of MorphOS. I doubt that we could find a group of ARM developers who could/would be willing to create an ARM version of MorphOS, based on the x64 version, when the x64 version is still being created, and the goals may not be set in stone yet. It may still be a moving target at this point in time.

To be clear, this is a question asking how many people would be willing to fund a bounty to port the x64 version of MorphOS to the Raspberry Pi 3 (or later), and/or the ARM architecture in general, after the MorphOS x64 version is completed (or at the same time the x64 version is being created, but only if another team of ARM developers could be found to do the work, and such work would not slow down the existing MorphOS Dev. Team, in their work to create MorphOS x64).

I am looking forward to seeing a version of MorphOS that is not hindered by the legacy of Amiga roadblocks, running on powerful, new x64 hardware, but if we could also have that same, unhindered MorphOS running on the Raspberry Pi 3, hopefully with full support for the on board GPU (which may be a problem), I would be even more excited to see what might be possible for potential growth of our user base.

I second the MorphOS bounty to The ARM Processor. I understand that ARM is much more difficult to port to, but Morph has always done the impossible before. One example is that they had no information about the AirPort cards, but they ended up supporting them.

Pi Device have many available OS on them, but if I understand correctly, you can get them with no OS pre-installed and install your own later, or select desired OS upon purchasing them. This is completely different compared with x86 Laptop, which can be purchased used with OS pre-installed onto them. Many people (probably almost all people) are accustomed to OS preinstalled onto computer and will not see the reason to replace the OS already existing on computer. They will usually just buy another computer with the OS (in this case, an updated version of Windows) and so if people do not like to update Windows, they will not want to install a "foreign" OS. I believe that people who use computers instead of upgrading existing computers would just get another computers based on their behaviour to get new Smart Phone every time one is released, are supposed to be "bigger and fancier". I do not care about having the newest SmartPhone, but I have found it difficult, or at least very annoying, to install an OS that already has an OS pre-installed. Additionally, I am not a "multiple boot" person.

Also I believe that there are many Laptops that are sold with Linux pre-installed, so MorphOS must have available computers to pre-install MorphOS onto ... or at least Blank computers with no OS that MorphOS can be installed.

Amiga Road Blocks: Yes, I believe that any old software needs to be updated and ported to the new Operating System / Hardware, or at least placed into separate emulator, instead of being a restrictive OS trying to be 68K software when it is actually PowerPC. Example: 68K Software should be made PowerPC, or placed in separate Emulator.

MorphOS Road Blocks: MorphOS would be nice to have Developer Tools. Pi Device could accomplish this.

Pi Device is very successful. If Commodore still exist today, they probably would want to use Pi Device.

MorphOS bounties are not really happy ending recently, there is no point on another one imho.

> they had no information about the AirPort cards, but they ended up supporting them.

That's because they finally looked at the available information :-)

> I believe that any old software needs to be updated and ported to the new
> Operating System / Hardware, or at least placed into separate emulator,
> instead of being a restrictive OS trying to be 68K software when it is
> actually PowerPC. Example: 68K Software should be made PowerPC, or placed
> in separate Emulator.

Are you confusing API and ISA here? I can't make head or tail of what you wrote here anyway.

@Incorrect,

I don't know why you are mentioning PPC software, as I thought I was perfectly clear that my proposal is to port the x64 version of MorphOS to the ARM architecture, and if possible, to the Raspberry Pi series of boards specifically. PPC would have nothing to do with the new version of MorphOS, unless the x64 version comes with some kind of PPC emulator, so it can run a few older, but still useful native MorphOS PPC applications.

Also, I don't know what you are talking about when you say "Pi device"? If you are referring to the tiny motherboards themselves, then there is nothing magical about a port to the Raspberry Pi series of boards that would make writing new software, or porting existing software easier. We would still need a new SDK for the ARM/Raspberry Pi version of MorphOS, and since MorphOS will be very different than Linux or Android, none of the existing development tools for those platforms will help us.

Still, I appreciate your support of this idea, as that is the only purpose of this thread, to gauge interest and begin discussion about the possibility of porting the x64 version of MorphOS to ARM/Raspberry Pi.

The x64 version of MorphOS is 1st priority, but after it is completed, hopefully we can convince the Dev. Team to port to the Raspberry Pi3 or later, with or without a bounty. I anticipate this to be a very long range goal, that will not be started for a couple of years or more.

You must not take things so literally. In that paragraph I was talking about Amiga Road Blocks. That software you and I mentioned is technically 68K software that MorphOS is designed to use even though MorphOS is not on 68K, but PowerPC. Basically the software designed to run on other hardware should either A: Be inside a separate emulator instead. and / or B: make a native version of the software that came from the other hardware.

Pi Device ... they make so many Pi Device ... Raspberry I, Raspberry II, Banana, etc. So many potential for name every time they release a new Pi Device. The Pi Foundations(s) think of so many names, Even Morph can have its own name for its very own Pi Device. Cherry Pi, Coconut Pi, Pecan Pi, Strawberry Rhubarb Pi, Lemon Pi, ... I could say BlackBerry Pi or Apple Pi, but those other companies may want them. Morph could in very distant future have its own Pi Device because there seems to be computers named after various fruits. Apple, BlackBerry, Raspberry, Banana are ones already taken. But I suppose I am being off topic. "Pi Device" I use as an umbrella term for any Pi Device. I think because of the nature of Pi Device, is supposed to be a "teaching how to code" computer, might attract developers to MorphOS.

Yes, I already agree with you there must be more expansions of Development Tools, and Software Development Kit. This will benefit any ARM version and also the x64 version.

X64 is next and main priority for Morph, so I agree if they proceed work on X64. If that means it takes very long time before starting ARM, this is very acceptable because I wish to donate to this bounty that you are proposing.

I cannot Lump Sum Donation to this bounty. So if bounty is open for a few years, I can donate small amounts which will accumulate to large amounts eventually. Then years from now if Morph is ready to proceed with ARM, then they will see the bounty and perhaps be inspired to expand to ARM.

The problem is that there are so many man-hours involved in porting to ARM (and making decent PPC->ARM and 68k->ARM interpreters/JITs) that a bounty just wouldn't cover it traditionally. You'd have to rely a group of people being motivated enough and free enough to do it. If the MorphOS team were that inclined you'd already have seen it happen, bounty or not.

Quote:

To the best of my knowledge there will not be any PPC->x64 or 68k->x64 translators either, other than what UAE has to offer. The x64 migration represents a clean cut from the lagacy compatibility, although the aim seems to be to keep portability as high as possible so that old MorphOS applications will require as few modifications at source code level as possible to run on the new platform. With that in mind, a second port to some popular ARM devices wouldn't be impossible at all IMHO, once the OS already has been adapted to x64. Whether it will make sense is a different matter. Porting it to x64 is porting it to some architecture(s), porting it to "ARM" is more about porting it to specific devices, due to the higher level of integration and differencies. It would also demand extra from SDK development to make supporting multiple platforms as transparent as possible for the application developers, ideally just setting a switch when building.

> Morph can have [...] its very own Pi Device.

Who should design and manufacture this? And why should an own device be necessary? And what or who is "Morph"?

> Morph could in very distant future have its own Pi Device because
> there seems to be computers named after various fruits.

The logic of this "conclusion" completely eludes me.

> "Pi Device" I use as an umbrella term for any Pi Device.

A self-referential definition doesn't make sense.

> the nature of Pi Device, is supposed to be a "teaching how to code" computer

A better term for this would be "hacker board".

http://hackerboards.com/raspberry-pi-3-takes-the-cake-in-2016-hacker-sbc-survey/

After all, most of these don't even have "Pi" as part of their name.

Quote:

It's really not that simple - not everyone used or uses the GCC compiler that comes with the MOS SDK. At very least, all of AmigaZeux's stuff for MOS was compiled on ECX, which was original 68k but was eventually adapted to generate PPC code, after many years. WOS/PUP compatibility is also important for many, and 68k is probably essential. If not, why not just use AROS - it's already on ARM and already doesn't run all your stuff better than you wouldn't run it on MOS :P

There's also the matter that, while the original (and now very underpowered) RPi is single core, the modern versions are multi-core, and most "competitive" ARM have four cores. A shared address space OS cannot ever take advantage of multiple cores.

@KennyR

I see your points, but perhaps you should read up on what has been said about the upcoming x64 port? AFAIK, decisions were years ago and work has been going on ever since...

At this point I don't care if they port it to ARM or X64, as long as it doesn't take forever.

> If not, why not just use AROS

Because AROS has the known limitations caused by the AmigaOS API compatibility.

> A shared address space OS cannot ever take advantage of multiple cores.

That's why MorphOS/x64 (and thus MorphOS/ARM as per amigadave's proposal) is supposed to be an SMP-capable OS.

> At this point I don't care if they port it to ARM or X64

amigadave's proposal isn't to port to ARM instead of or before x64.

Quote:

but at least it's the most hilarious stuff i've read from here in a long time

Quote:

As TMHG wrote, you need to read up on what decisions have been made with regards to the x64 version of MorphOS being a clean break from Amiga 68k legacy limitations. AFAIK, MorphOS x64 will be multi-core, SMP and also include memory protection. Maybe it will also include multi-user support, but I doubt that is a high priority. At one time, I proposed that the MorphOS Dev. Team join with AROS developers, once they decided to move to x86/x64 and/or ARM, since AROS has been working on these platforms for quite some time, but I now see that the goals of the MorphOS Dev. Team are much higher/better than the AROS goal to duplicate the AmigaOS 3.x API, without removing the limitations of that legacy OS, and providing modern features needed by today's computer users. Perhaps there is still an opportunity for MorphOS Dev. Team members and AROS developers to work together, for the benefit of both, as they have done in the past. There is most likely some work already done on AROS for x64 (and for my proposal, AROS for ARM), which could be useful to the MorphOS Dev. Team, and certainly new work done now and in the future on MorphOS x64, will be useful for the advancement of AROS on x64, but since one platform is closed source and commercial, and the other is open source and free, there will be only a limited amount of sharing going on.

If current applications for MorphOS3.9 were not written with future portability to another architecture in mind, then it is the problem of the programmer, not the MorphOS Dev. Team members, to get those applications ported to the x64 version of MorphOS. I think that the MorphOS Dev. Team members know that availability of new, or ported software for the x64 version of MorphOS will be important for the release, so they will do everything they can to create good tools to make porting or existing and creation of new software, as easy as possible, so we will soon have many applications and games to run on MorphOS x64, after it is released. As I wrote earlier, all 68k & maybe even PPC software, should only be run via emulators, like UAE, under the x64 version of MorphOS, with no compromises made during the creation of MorphOS for x64.

I hope that one of the goals of the MorphOS Dev. Team members, is to keep MorphOS as small and efficient as possible, so a possible future port to low power ARM devices will be feasible. This whole thread will be moot, if the x64 version of MorphOS becomes bloated and will not run as fast, or faster, than Linux & Android, when/if ported to ARM.

Quote:

"or even Linux"... kinda seems like Linux is somewhat within grasp of MorphOS. Far from it. LOL. Dont make MorphOS to be a competitor on any level against any OS other than "Amiga like" operating systems. It will lose.
Having more hardware support is great, its not ready to fight against a very very established OS like Linux.

Quote:

Quote:

I did not know of the umbrella term "Hacker Board" and only thought of "Pi Device" based on success of Raspberry Pi Foundation and other related device. Example: "Pi Device" umbrella term I thought much better to say instead of "Raspberry Pi, Raspberry Pi II, Raspberry Pi III, Banana Pi, and BeagleBoard." There are so many "Pi Device" that they even creating the Raspberry Pi III before I noticed! (However, it is not surprising that they are making the Raspberry Pi III as Pi Devices are very successful.)

Quote:

and

Quote:

Morph, Apple, Blackberry are free to select their own custom Pi Device if they feel the need to have their own Pi Device. Or any Hacker Board, but in the case of Apple and BlackBerry they may wish to choose corresponding fruits, such as "Apple Pi" and "BlackBerry Pi", respectively. Morph can select any other available "Pi" name, or they can name their own Hacker Board. Morph having its own custom Hacker Board could give MorphOS more recognition. Or perhaps if may never make sense to do so. In that case, MorphOS can be run on any Hacker Board but which one to pick first??


Quote:

MOS was a company that manufactured Commodore Business Machines hardware. Neither companies today exist.



[ Edited by In_Correct 27.07.2016 - 23:24 ]

Quote:

MorphOS is not ready to compete from Linux. I believe everybody is talking about a possible distant future when MorphOS has been ported to X86 and then to ARM, with Software Development Kit, Developer Tools, and Many Software. It can then have a chance to compete with Linux.

Linux has become a leader in the current computing marketplace. This can change if there was a new target audience marketplace, such as in Africa.

> Morph [...] are free to select their own custom Pi Device

How does this "selection" process work? Who will design this device? Who will manufacture it? Who will pay for all this?

> if they feel the need to have their own Pi Device. [...]
> Morph [...] can name their own Hacker Board.

Why should they? Wouldn't it be better to select an already existing and successful board for a bigger potential install base?

> Morph having its own custom Hacker Board could give MorphOS more recognition.

I doubt this would work or make economic sense.

> MorphOS can be run on any Hacker Board but which one to pick first??

The AArch64-based board with the biggest existing install base, I'd say.

Quote:

This all sounds like Hyperion's MAP -- pipedreams and fairies.

It also ignores completely the core dogma of OS usage -- people don't use an OS to use an OS, they use an OS to use software. OS4 forgot this and made an expensive platform almost none of its original fanbase ended up using, simply because there was nothing to run on it. (Well, not completely true -- OS4 won a few converts over MOS because its support for AGA and the native chipset.) The world is replete with good operating systems like BeOS/Zeta that went nowhere because there just wasn't the userbase, software, or drivers.

I don't know about y'all, but I used MorphOS because it finally let me use my Amiga applications on a system that wasn't designed in the 1980s and on a CPU that had enough power to do basic computing in the early 21st century. I didn't use it because it wasn't Windows or Linux -- Windows and Linux, which are fully SMP-capable, x64/ARM operating systems, work just fine for me.

> MorphOS is not ready to compete from Linux.

Competing or not... There are many aspects if comparing two OS, but wo cares exactly?
I use MOS on every day basis without SMP and Memopry protection. With these features, MOS will be more modern and (probably) reliable system. Hopefully, the supported HW will be as lovely as my PowerBook is.

Quote:

Quote:
I hope that my recommending a "Cherry Pi", "Lemon Pi", "Coconut Pi", "Pecan Pi", and "Strawberry Rhubarb Pi" does not sound like Pipe Dreams And Fairies. It is not my intention.

Quote:
This could and should be accomplished with emulator or otherwise the software updated properly.

Quote:
This is why I am here. I despise Windows and Linux. For one useful feature that they add, they add twenty or thirty useless features. While it is possible to undo the useless features, the adventure is extremely painstaking. I am very positive with this new MorphOS because it will have the same modern capabilities SMP 64 Bit With Memory Protection but focusing on performance. The direction that other OS (and applications "apps" and websites) are taking is that they make computers toys, not tools. I do not like it when "user friendliness" treats me like a small child. The "Oops!" error messages are the perfect example. The message says "Oops" implies that errors are acceptable, or that I made the error. It is not acceptable; and I did not make the error.

Quote:

IMO, if I was the Dev Team, I'd stop at x86 unless they're bored. A wide range of new software is what MorphOS needs. Not current re-treads of 68k software. Hopefully with new hardware that'll happen.