I know we've been over this before and everyone has pointed out that since the original AmigaOS had no support for multiprocessing, Abox cannot support more than one processor (or core).
I understand the arguments about compatibility.
But what about the Quark kernal, can it handle a multiprocessor environment and can it support the concurrent operation of the Abox with other processes?
I know the Qbox has essentially been abandoned, but it seems like another process (possibly with less compatibility than Abox) could be run alongside our current operating environment to expand it.

As far as I know, MorphOS consists in a single process (the ABox) ran by Quark, which has everything inside. I guess Quark can have more processes, but what would you do with those? You're talking about writing a completely new operating system here.

You haven't spelled "kernal" that way by mistake, did you...

Quote:

Yes, if anything, I am the King of misspelling.

As to a completely new operating system, I'm not sugesting that (yet).

But you have pointed out that MorphOS (and Abox) are self contained. Quark, as a microkernel, probably only offers a few routines that could be accessed by MorphOS (possibly none).

There isn't any documentation on Quark (that I can find), but it's not unreasonable to expect it to be able to control timeslicing (if not SMP).

I guess what I'm suggesting, at this point, is not a new OS. Rather that features that exist in Quark now (or those that could be added) might help support SMP transparently to MorphOS.

I'm not that intimidated by the idea of a new OS, if the kernel is already written. The first computers I used had no real OS, just a bootloader. Assembly language for a PPC looks a lot easier than an X86 (which, by now, has far too many instructions to want to program w/o a higher level programming tool).

If there was even modest documentation on Quark (enough to initiate another process - or start to create the modules necessary to support additional processes) I'd be willing to devote some time to working on this idea.

That's the genius of micro kernel based OS'. Only what is absolutely needed is present in the kernel, but additions can be made to add functionality.

[ Edited by Jim on 2010/4/23 20:27 ]

> what about the Quark kernal, can it handle a multiprocessor environment and can
> it support the concurrent operation of the Abox with other processes?

My stance on this, which you know already:

"Quark probably has been supporting multi-processing from the beginning."
https://morph.zone/modules/newbb_plus/viewtopic.php?topic_id=6267&forum=11#62573

More threads on this subject:

https://morph.zone/modules/newbb_plus/viewtopic.php?topic_id=6248&forum=3
https://morph.zone/modules/newbb_plus/viewtopic.php?topic_id=6319&forum=12
https://morph.zone/modules/newbb_plus/viewtopic.php?topic_id=6570&forum=3

I remember most of this Andreas.

And I can understand the development team's focus.

I just don't see why further work couldn't be done outside of MorphOS without breaking the Abox.

On faster processors, it should be possible to do this transparently (w/o significant modification to the Abox).

And yes, we have covered this before. But what other options will be possible (for future advancement)? Especially if AmigaOS somehow evolves to support SMP.

Sorry, I've apparently pushed too hard on a topic that is a sore point with a few of you.

It's just frustrating knowing that there were originally plans for so much more and now you've settled for doing a really good job on what originally would have been the first step in the evolution of something more...

A quick question. Who created Quark?

> Who created Quark?

https://morph.zone/modules/newbb_plus/viewtopic.php?topic_id=6248&forum=3&start=20#62268

MorphOS key feature, from the start, was its ability to be a succesor to AmigaOS. Alright, it might be based on something entirely unrelated (Quark), but its use from the beginning was that. If it wasn't, then it would have been just another experimental operating system with zero applications (and drivers!) tu run on.
So the choice was right, it was either that, or an absolute zero starting point. And of course, all of its "fathers" came from Amiga land.
So I guess MorphOS role in the world came out of mere circumnstances: Sure, it's based in a concept where you could have several "boxes" for different systems, but it has never had more than one - and very well done. There was no interest in doing something else with Quark.

Small correction: There may have been interest all along, but the limited development resources have been focused on other issues with a higher urgency as you explained.

Thankyou Andre. Could you tell us about those different interests? Who else wanted MorphOS, aside from aimgans?

Actually, I could see where MorphOS might succeed in being BETTER than AmigaOS (with features amigaOS lacks) and garner attention outside the Amiga community.

Quote:

...and you won't tell us either?

@ jcmarcos

I was talking about the motivation and interests of (some) developers, not actual or potential users. If "doing something with Quark" referred to users for some reason, I think there are sufficient discussion threads on this very website that show a certain level of interest.

Unfortunately, "interest" does not necessarily translate into anything tangible. A person might have some general interest in climbing the Mt. Everest, but that does not mean he will in fact do it :)

Actually, that's the reason behind my post.

Personally, I've never understood people foolish enough to risk their lives doing something as dangerous as climbing Mt. Everest.

But wanting to learn more about the potential of (what looks like) a very powerful micro kernel? Yes, I'd want to do that, paricularly if Quark could spawn other processes that could run concurrently with the Abox.

I never suggested a return to the development of the Qbox, just exploiting features that might be present in the kernel.

Jim has a plan! He wants to use a multiple processor computer, running one Quark on each processor, with one of those also having MorphOS inside. Then, from there, he would launch Quark tasks in the other processors, just preparing world domination! Muahahahaha...

Sarcastic, but not amusing. Not accurate either. I'd like to kmnow how well Quark supports time slicing and what other capabilities might be support in the kernel. I'll leave world domination plans to the US Republican party.
I don't know that a second copy of Quark would be necessary to spawn additional processes (outside of the Abox).

But without documentation, this argument is a moot point.
However, I don't think its funny.