(Thread clean up.)

Fan control is outside of Thermals app. That's just a frontend to display stuff.

Either way, this has nothing to do with Wayfarer.

[ Edited by jacadcaps 09.09.2026 - 00:28 ]

sorry, but I’m not sure which one you’re referring to
If you mean ‘Thermals’ in the Utilities folder, when I run it on my PMacG5 11.2 I get the message: ‘This architecture is not currently supported by Thermals’.
What should I use to view and adjust the fan speed settings?

I think you're confused. That tool you've linked to is obsolete. MorphOS has proper fan control code for all of the models you've mentioned.

Quote:

Great news, Jacek!
I have a request: since this will also tax the cores or additional CPUs depending on the different PMacG5 models, it would be helpful to have a utility to increase the fan speeds on these Mac models as needed, something similar to this software https://crystalidea.com/macs-fan-control/download or this one https://www.morphos-storage.net/?id=1604130 which would allow for better management of everything via a GUI
Incidentally, unlike the 7.3 and 7.2 models, the 11.2 models have a shell command for fan control that isn’t usable in MorphOS (if I remember correctly, this command is called PowerMac7FanControl), so it would be very useful to have something that works for all PMacG5 models and even better if everything could be adjusted via a GUI
What do you think?

> More in due time. I don't have all of the hw you've listed,
> so it's unlikely all of those will be supported.

Fair enough. In case of support for Mirari's 3 additional cores, is one video decoded by all 3 cores in parallel (= triple decoding speed and thus ability to decode videos one core would be too slow for), or still just one core per video (= up to 3 videos in parallel)?

> I thought using the additional core(s)/CPU(s) was not possible
> with Amiga type systems?

What's not possible with Amiga-type systems (while staying compatible) is symmetric multiprocessing. In contrast, what's been done here is asymmetric multiprocessing (as linked to in comment #1927) and has been discussed for ages on MorphZone as a compatible way of leveraging additional CPUs/cores. Random examples:

https://morph.zone/modules/newbb_plus/viewtopic.php?topic_id=7405&forum=3
https://morph.zone/modules/newbb_plus/viewtopic.php?topic_id=8096&forum=3
https://morph.zone/modules/newbb_plus/viewtopic.php?topic_id=8831&forum=11
https://morph.zone/modules/newbb_plus/viewtopic.php?topic_id=9204&forum=11
https://morph.zone/modules/newbb_plus/viewtopic.php?topic_id=10057&forum=16

Similar ideas (in broad concept, not necessarily implementation-wise) are old PowerUP/WarpUP and recent XACP with the difference that what MorphOS 3.22 is supposed to bring uses coprocessors of the same ISA instead of different ISA, which (in addition to hardware cache coherency) may have simplified things in comparison.

Kronos broke the internet. Again!

@NewSense

Pretty sure there will be a 3.21 release focussing on bugfixes and a 3.22 with new features some time later. At least thats how it has been done in the past and is supposed to be done.
Timeline for both is anybody's guess as usual.

These extra tasks will run outside of ABox, so no (direct) calls to legacy API no way to run existing apps this way (obviously) but also no restrictions imposed 40 years ago.
What isn't clear whether these "QBox" ( *runs*) tasks are still 32Bit or 64 which would completely rule out anything G4.
Not that Dual_G4 would make much sense in the 1st place with the limited number of people running these and the hard limitation to either 2 or 1.5GB.

[ Edited by Kronos 06.09.2026 - 04:52 ]

Quote:So, will there be a v3.21 to fix the demuxer issue, or will the MOS Dev. Team be going straight to v3.22?
Quote:So, does that mean that the tasks to action the video, audio, image decoding are run as separate threads/tasks, or is it implemented in a different way?
I thought using the additional core(s)/CPU(s) was not possible with Amiga type systems?
Is this your, or someone else's idea to implement this functionality?
What speed increase is there to the system's CPU(s)/Core(s) when this method is implemented?

More in due time. I don't have all of the hw you've listed, so it's unlikely all of those will be supported.

> Dual core/dual CPU utilized in G5 Macs, 4 cores on Mirari.
> Video, audio, image decoding offloaded to the additional cores.

Wow, that's insane :-D Must be a decade or so since I heard of bigfoot's successful attempts at running something tangible on further cores AMP/ASMP style.
What about the 2nd CPU in PowerMac Dual-G4, or 3rd and 4th cores (2nd CPU) in Quad-G5, or 2nd to 4th core in X5000/40?

Quote:

Bamm!

Deeply impressed. As said, still happy about the JIT speedup and now a steady supply of groundbreaking improvements. Wow, wow, wow!

Coming with MorphOS 3.22:

Youtube

Dual core/dual CPU utilized in G5 Macs, 4 cores on Mirari. Video, audio, image decoding offloaded to the additional cores.

My jaw is on the floor. I can't wait to test this out.

Quote:

Perfectly Said! Love it...

I might ask to borrow it sometime.

@ jacadcaps - Most impressive,

. . . . . Most impressive INDEED . . . Young Skia-Walker

So, seems it does, that a great new browser of the web built have you! young Jaca - wan in a million you seem to be!

Quote:

Yes, the Skia-CPU path has been extensively tested and got a lot of Altivec love too. That said, the difference is well worth a hw upgrade.

Quote:

I praise you! Still happy since the JIT update, which gave the crucical speed boost to use Wayfarer as daily driver again (for most of my needs!). Another speed bump is always welcome. Great!

Quote:
Yes it will.

That's great! :) But Wayfarer 12 will still work on R200 cards, without GL rendering?

Wow, hw rendering, that’s great!