The MorphOS development team is proud to announce the public release of MorphOS 3.5, which introduces support for PowerMac 7,2 machines and features various bug fixes as well as other improvements. For an overview of the included changes, please read our release notes.
They released 3.5.1 shortly after to fix a boot issue in 3.5.
Should be;
http://www.morphos-team.net/releasenotes/3.5
I’m disappointed they still don’t support G3 Macs. They have G3 support for other systems, and the G3 Macs are the same as the G4 (except for the occasional newer video chip), so why is there no G3 Mac support? I’d buy MorphOS for my iMac G3 in an instant if it worked.
Yes, that is weird as MorphOS worked on the EFIKA that was a G2/G3 bastard without AltiVec. You still needed a Radeon card though.
Pretty simple: supporting G3 Macs takes time and resources better spent on higher performing machines.
IIRC the G3 Macs hardware varies much more than the G4 machines which would make it harder to support too.
The G3 iMac was the biggest seller of the time, and all models are damn near identical other than speed and amount of ram. They did a couple updates of the video chip, but most are ATI Rage. Of all G3s, it would be the easiest to support, and give a HUGE boost to the number of Macs covered. And given they DO support a couple G3s with VERY “odd” architectures, claiming the iMac G3 isn’t worth supporting is bogus just on the face of it.
… keeps getting better. It’s just a shame that the dev team has hitched its wagon so firmly to old PPC Macs. At the moment, the hardware is still cheap and plentiful on eBay and the like, but the time will come when that’s no longer the case and I wonder what will happen then.
I’m not sure what PPC hardware they’d target other than that though. Only other PPC based systems out there are the IBM iSeries and of course the AmigaOne X1000, but I think most people buying those are running AmigaOS4 on it.
Personally I think AROS has the most future, since it’s actually open source and runs on other operating systems.
It’s always saddened me how fragmented the Amiga community is. I swear I want to invent a stasis bubble for my Amiga 4000D so that it never goes bad on me.
How about the fastest, cheapest, best selling PPC platforms of all time? Namely, the PS3 and XBox 360.
well those systems are pretty locked down and would be tricky to port to, but there is a bigger problem: MorphOS cannot handle more than one core. So MorphOS would be running on a single non superscalar core on the 360, for example.
But the problem really remains the same, its just running on legacy hardware. The PS3 and 360 would be manufactured for a little bit more, but the best by date is rapidly approaching.
If MorphOS wants to do anything but stave off the merciful release of the reaper, they have to get onto hardware that is actually being manufactured.
But to paraphrase Douglas Adams: It seems like they knew all about survival of the species, but wanted no part of it…
Yes, the multi-core nature is an issue, but not a huge one. Not these days. At the very worst, they could simply not use the extra cores (or the hyper thread in the case of the PS3). Both the PS3 and XBox 360 have exploits for running homebrew, but MorphOS should have the creds to actually be a licensed producer for either or both platforms.
You’re also right about the need to work on something that is still going to be manufactured in the future, which basically means AMD64 or ARM platform devices. But in the meantime, since they like targeting PPC platforms, the PS3 and XB360 gives them tens of millions more to target. Just like I’d buy iMac G3 MorphOS, I’d buy PS3 MorphOS in an instant.
Edited 2014-02-19 23:53 UTC
Let me suggest a different argument: It would be non trivial to port to either of those game consoles. It would be easier to port to the G3 Macs people have been discussing than to the XBox 360 (which should be the easier of the two platforms).
Would that be worth the effort? Would enough people move to MorphOS 360 to pay for it.
I say no. My best guess is that the MorphOS market is is pretty saturated. Its only real hope is to move to a platform where people would be interested in in an alternative OS.
There are versions of Linux for the PS3 and 360 but they aren’t widely used at all. People aren’t really clamoring for it. And my suspicion [Game Console] MorphOS wouldn’t attract many new users, it would just shuffle existing customers.
However I think MorphOS would have a real chance of growing its userbase if it were available for the Raspberry Pi (or better yet the BeagleBone Black). Don’t worry about PowerPC/68K compatibility and encourage recompilation for the platform.
Price it at something much more reasonable ^aEUR” 20 EUR ^aEUR” and I think it would have a chance of getting a growing pool of users.
But I don’t see that happening. From what I can tell they are more likely to do a x86 port, and they just don’t have as much to offer there (IMHO).
I don’t think Amigans ever liked PPC in itself, it was more about an irrational hatred of x86 and hence liking any alternative to it.
Hating x86 isn’t irrational, it’s just an indication of good solid understanding of the technical nature of CPU architecture. EVERYONE who doesn’t work for Intel who understands CPUs hates the x86. I never knew another engineer who liked programming the x86 in assembly. People LOVED programming in assembly on the 680×0… and the PowerPC. That’s why Amigans liked the PPC. At least, Amiga PROGRAMMERS did.
I do have to give the x86 one bit of credit – if it hadn’t been such a disgusting piece of sh-t, compilers would have never have advanced as quickly as they did. Programmers hated x86 assembly with such a passion that they quickly improved compilers to a state where they no longer needed to use that hated assembly ever again.
The best thing you can say about AMD64 is that it makes the x86 (in 64 bit mode, at least) very UNLIKE the x86. It smooths out the worst of the x86 features, and adds in features to make it more like better ISAs.
And I meant mostly Amiga users, not programmers. The former also very much seemed to hate x86, even though they could only benefit from it. Recall Amithlon (a hypervisor of sorts, which presented a fairly standard PC as an Amiga to the OS), which was by far the fastest Amiga at its time (and probably still would be, if it haven’t been killed, if it supported latest PCs)
BTW, the CEO of Apple during 68k->PPC transition has stated that going PPC and not x86 was his biggest mistake…
It makes us programmers sad that x86 “won” the CPU wars.
It’s partly why my hobby is programming old consoles like the Genesis. I can do some good old-fashioned 68000 assembly, and people still appreciate it.
The truly funny thing is that PPC would probably occupy the same position that ARM does today, if only it hadn’t been hobbled by guilt-by-association with over-priced, under-performing Apple products.
I am sorry, but you are going to have to back that assertion up with some facts. At the time Apple’s products were the main thing keeping PPC going, outside of the the high end and low end. Nor was ARM hurt by its bad representation in the Newton.
I do embedded work (along with other kinds of engineering) and do you what I have never heard a embedded engineer say? “Oh I would never use PowerPC because of Apple”.
In my experience PowerPC is used quite a bit by the embedded crowd, mostly via Xilinx FPGAs but there are others. However it is almost always at the higher end of embedded (again in my experience). I don’t think I have ever seen one that ran on batteries.
And that right there is why you ARM, AVR, PIC, MIPS, 8051, MSP430, et al all get used more than PowerPC microcontrollers: Power Consumption. PowerPC has great {Insert Computing Unit of Choice} per Watt at the high end, but not at the low end. Even the Sparc microcontrollers (in the form of LEON cores) are leaner.
Only P.A Semi (now owned by Apple) with their PWRficient made good low power PowerPC cores (still not as low as the others I mentions, but not bad). They are committed to producing them for a couple more years to fulfill military contracts, but they likely won’t be around much longer.
Yeah, except for a little company called “IBM”. Perhaps you’ve heard of them?
And don’t bother pointing to the “AIM” alliance, it was always obvious that Apple was the third-wheel of THAT trio – it’s the equivalent of a superhero team with a roster consisting of Iron Man, The Hulk, and…. Paul Blart, Mall Cop.
Maybe that’s because there were numerous other consumer products using ARM, preventing it from being solely-associated with Apple in the minds of end-users…. just a thought.
And that just reinforces my point: PPC is a good enough arch that even Apple’s negative halo couldn’t kill it off entirely.
Read what I wrote again. I said Apple kept PowerPC alive in the middle. IBM keeps it alive in the high end, but has nothing in the middle or low end.
Can you please point to ANY PowerPC presence IBM has in desktops or workstation type systems? Because I think you are going to have a very hard time doing so.
My point still stands.
Um no. Your point was this:
And I asked you for data to prove that. Which you haven’t done, btw. I very much doubt you can, but please try. But if you argument is simple “see they use it in embedded, it must be good”; then I will invite you to look at all the weird oddball architectures still alive in embedded.
The single most popular embedded core in the embedded world is the 8051. I know no one who sings its praises, except for the fact that it is used, well supported, and more or less free. So much research has been put into making it fast and cheap that is it more or less self perpetuating. But no one really likes the architecture.
The 6800 and 6809 still live in embedded, but that is because of legacy and not because of some proof of the architecture.
I also pointed out the main reason why PowerPC doesn’t have wider adoption in embedded: power usage. The reason PowerPC isn’t more widespread is because IBM and Freescale never emphasized it. Xilinx did more to popularize Power in embedded than IBM or Freescale combined.
Why would I? Those three companies were effectively the only companies to do anything with the Power Architecture. IBM and Motorola (now Freescale) failed to keep the architecture competitive, IBM and Apple were the largest customers… until the Game Cube/Wii/Wii U, PS4, and XBox 360 came along. The Amiga market was so small as to be considered line noise compared to all of the others.
In my opinion Apple didn’t hold the alliance back, the alliance held itself. back.
I still fail to see how any of this has anything to do with Apple ^aEUR” except for your apparent desire to blame them. In my opinion, you don’t seem to know much about the embedded world at all.
Unless you can come back with some verifiable proof that Apple was responsible for stifling PowerPC; I have no reason to engage with you
Doesn’t IBM have some middle-class PowerPC cores used in massively parallel systems?
I wasn’t talking about the cores, but the systems. I can’t name anyone who is making PowerPC desktops, laptops, or workstations… aside from the AmigaOne X1000; and I wouldn’t call that a particularly vibrant community.
Indeed individual PowerPC cores don’t stack up that well against x86; IBMs advantage with Power is in performance per watt and number of cores per chip.
I wanted to address your Newton/ARM halo comment:
Seperately. At the time Apple introduced the Newton (and for some years after) the only other ARM products on the market were from Acorn (ie the RISCOS stuff) and essentially was unknown anywhere outside of the UK/Europe.
ARM Holdings aggressively marketed the ARM core and licensed it everywhere they could.; but it really wasn’t until DEC evolved it into a superscalar architecture ^aEUR” ie the StrongARM ^aEUR” that it really started to catch on.
There is a lot more to that story. The rise of ARM, the decline of MIPS (still very much alive in embedded, far moreso than PowerPC), and finally the RISCification of the x86 architecture. It was a really exciting time for the CPU nerds.
My real point here is to point out why ARM (and MIPS to a lesser degree) thrived when PowerPC merely has a place at the table: ARM and MIPs licensed their designs far and wide. They let others improve them, and brought these improvements back into the core design. They evolved far more quickly then the Power Architecture did.
Why? Well I believe it was because IBM wanted a top down process. They designed the cores, Motorola/Freescale manufactured them, and Apple & IBM used them.
IBM couldn’t compete with the race to ever smaller die sizes, and Power was a component heavy design ^aEUR” ARM was quite lean by comparison. ARM manufacturers weren’t limited to Motorola and IBMs fabs; ARM allow for higher yields per die (which means prices were lower and required less power).
You could see this as an example of the Cathedral vs the Bazaar; or of anything you like. But in the end ARM won over PowerPC because of its merits.
I wonder what the MorphOS team has in store for the future, with PPC dying as a desktop computing architecture. It’s a clever little OS, but sound engineering alone isn’t helpful without a strong business case behind the product.
My understanding of MorphOS is that it currently runs the “ABox,” compatible with the Amiga API’s, on top of the Quark microkernel. The currently-undeveloped “QBox” would be capable of modern OS functions like symmetric multiprocessing, but would not be Amiga-compatible. The QBox does hold potential for the future; if there’s any room for another competitor in the world of RealTime Operating Systems, MorphOS would be wise to recast itself as an OS for the high-end embedded PPC boards, particularly rad-hardened spacecraft avionics boxes.
While there’s been talk of rewriting and recompiling MorphOS to run on ARM, it’s unlikely to achieve any significant market share in that arena due to the phenominal market performance of Android. It’s conceivable that in the same way MorphOS has given new life to the Macs that Apple abandoned, an ARM port of MorphOS could give new life to tablets abandoned by Asus, Samsung and the rest (albeit with much more work required to support the multiple tablet configurations that have been produced so far.)
There are so many (free and commercial) RTOSs out there, I doubt that MorphRTOS could make much of a dent. Never say never, but I don’t see it happening; or what they would bring to the table.
As for other architectures I would imagine they would handle them the way to do 68K code right now: Emulation & Recompilation. If they ported to ARM or x86 then they could introduce PowerPC emulator(s), and the more active projects could recompile.
I suspect they are more likely to port to x86 and be a niche os on that architecture. I recall reading a thread where the MorphOS team worried that ARM might just be another fad like PowerPC.
But let me play devil’s advocate for a moment: Why should MorphOS survive? Aside from letting you run legacy AmigaOS apps on legacy hardware; what does it offer?
Of the three amiga operating systems (AmigaOS, MorphOS, and AROS), only AROS actually planned on being portable. After one shift of architecture you would think that both AmigaOS and MorphOS would designed/refactored their codebase to make it portable. To the best of my knowledge, neither did.
Now still with my Devil’s Advocate hat on: What does MorphOS bring to the table if it doesn’t have a body of Amiga Apps to run? Very few of those will have a decent UX on a touch tablet.
Apple was one of ARM founders. They have top ARM processor design team. ARM should be doomed right now.