This is a table that was posted as an article by kallisti5. I have moved it here because as an article it is incomplete, but the information can still be useful to others.
Date
Machine
Processor
Memory
Revision
Status
Notes
11/2009
Apple QuickSilver
PowerPC G4 (7450) 867Mhz
1GB
r33991
Bootloader boots, finds Haiku root partition on USB stick, jump into kernel fails
-
11/2009
Apple PowerBook Bronze
PowerPC G3 (750) 333Mhz
192MB
r33991
Bootloader boots, finds root partition on CD, jump into kernel fails
Bootloader does not see USB drives.
11/2009
qemu-system-ppc
qemu emulated cpu
256MB
r33991
Bootloader will boot with modifications but not draw menu properly
openbios-ppc does not support the needed functionality to boot at the moment. OF /cpus/ is not implemented well enough by openbios.
11/2009
gxemul -e g3
gxemul emulated G3 cpu
256MB
r33991
Bootloader will not boot.
OF is not emulated well enough to boot at the moment.
Hi,
There is very slow progress on getting it to build again.
given that Apple machines with a PowerPC are not that common anymore, the current target is the Sam440 board which usually runs some variant of AmigaOS.
I have to respectfully disagree about suitable PPC Macintosh hardware being hard to find. MorphOS has supported various G4 (& even G5) ATI equipped Macintosh models and it’s been great for that community, who typically source through ebay or similar sites. I’ve got a G4 Mac Mini that I’ve been using quite regularly for some time and it’s never skipped a beat - same with my G4 Powerbook.
Please feel free to download and install the existing PPC ISO, do some testing, and provide a write-up and bug reports. More people using it is the best way to get interest.
Resurrecting an old thread. I know that interest in PowerPC is pretty low in general and I don’t know how far the current x86 nightly images has come compared to the old PowerPC ones code compatible wise (as in compiling the current sources with a PowerPC target), but the new Power9 based Talos II made by Raptor Engineering seems like a good target machine for Haiku PowerPC. Given that someone interested in PowerPC, Haiku and being able to code in C++ actually can afford to buy one ($4k USD). Still, it would be cool to see Haiku running on a 44 core IBM Power9 system.
It is formidable that they managed to keep going and are now able to design and market a second iteration.
Interestingly, there is a project for a PowerPC notebook which seems to be progressing well having raised sufficient funding for the developing the schematics - a 12,600€ scope of work ( https://www.powerpc-notebook.org/en/ ).
I don’t know how many developers have experience with the PowerPC platform. Nevertheless, it could be exciting if, one day, one of these systems gets loaned to Haiku for putting together a PowerPC variant of Haiku.
I’m following the PowerPC laptop project eagerly. I really hope it ends up developing a real product released to the masses.
I’m saving up for a Talos II system. Haven’t bought a new-new computer system in years, so I think I’m due for one
Well, there’s at least a handful of people working on the PowerPC targets of various Linux distros. There’s FreeBSD and NetBSD for PowerPC as well.
That’s good that the “BSD” projects are also aiming at PowerPC targets. Maybe some of the code, notably the device drivers, would be easily portable to Haiku.
I am also interested in an PPC port… not because of newer hardware but because of old Apple PPC hardware. A lot of this stuff ist still out there running. As Haiku runs quite nice on older x86 hardware it would be great to resurrect iBooks or G4 Cubes with Haiku. I could test Haiku on an iBook G4, a G4 Cube and on a PowerMac G5.
The old Mac OS version run quiet nicely on these systems, but browsing the web is a pain… Being not able to use a modern browser deems these machines useless. Using a modern Linux like Debian is possible, but the software wants more power. Using firefox on a iBook with Mint or Debian is no fun.
Haiku would fit in this niche nicely.
As far as I’m aware, the main appeal of Haiku is that it’s a direct BeOS successor. Because of this, I humbly think it’s a scope-level mistake not to target the PowerPC platform, as that is the main BeOS architecture throughout its lifetime (and the only architecture for half of its lifetime), with even the incomplete, unreleased 5.1 Dano version having a lesser-known PPC version.
I personally only got to even learn of Haiku because of its PPC roots, and I’m definitely far from being the only one. With G3~G5 Macs being easily and globally available at dirt-cheap prices (just got a G5 Quad and 1.5GHz mini, $200 and $50 respectively, and I don’t even live in the US or Europe, where availability is even higher), and with extremely high-end machines like the Talos II, Talos II Lite and soon Blackbird available for ordering and using today, I really think there is no better time than now to stay keen and invest on PPC, especially considering general system security (x86 and AMD64 processors have unremovable processor-level backdoors, lest you render the processors useless. For those interested, see Intel ME and AMD PSP).
I’m willing to test PPC Haiku on those machines, but download links to such Haiku builds have long been removed. I see there is still a download page, but with no downloads available.
In any case, please reconsider PowerPC for Haiku. Thank you for your time and for listening.
I think you’re missing the point. We are a small team who work on Haiku in our free time. There’s quite a lot of things we simply can’t do (or can’t do very quickly) as a result. That’s why there isn’t a fully working build of Haiku PPC already.
If I may ask, the point of what, exactly? Haiku’s scope? The dev team’s wishes of developing any non-Linux non-BSD OS as a hobby, regardless of its relation to BeOS? (Also, I’m a bit confused, but aren’t there hired developers, as well? Haiku Inc. is a registered company and said to own the rights to Haiku.)
I did mention I’m willing to test PPC builds, but never did I say or imply anything to rush, force etc. any of you guys, to warrant that reply approach.
The whole post is about Haiku’s stated goals (scope), and why PowerPC could be more central to it. From Haiku’s page: “[…] x86 is the only CPU architecture being targeted for R1.”
Non-x86 builds are being treated as, at best, second-class citizens. Still, there used to be PowerPC nightly builds, but in recent years, even those are now gone, while even Arm has some still up. Even the build page incorrectly and vaguely states “large amount of reliability issues with Apple PowerPC hardware and the large hardware cost”, wrong on both accounts.
So as I see it, the path Haiku is going is further and further away from PPC, and that now would be the one and only time to raise this issue. The issue has nothing to do with developer time, but with what is done in that time. Again, this is not meant to rush nor force any developer into anything, but to highlight the perception that PowerPC is being abandoned by Haiku, be the hardware new or old.
