Mainstream HW vs. Custom Niche HW

[Crumb]

I guess the question about G5's is... it'll give some more machines to use but what can you actually do on "amiga" that requires that level of hardware?

Encode audio/video, render graphics, play games, run emulators, compile big projects...

I would prefer that MorphOS was firstly ported to G5, multicore & 64bits support could be added later and in the switch to 64bits work could start in compiling everything for boring little-endian architectures. MorphOS uses IPTR pointers like AROS in its sourcecode and drivers apart it shouldn't take 5 years to port it to another architecture if you lose Trance, it could start its life internally in x86-64 hosted on another OS just like AROS did to avoid spending too much time on drivers for hardware that won't be available in the final release, and when it's more or less stable then focus on drivers.

As Iggy has pointed out the main problem would be lack of software, a ppc emulation box containing a full PPC-MorphOS could be used for that, there are several opensource ppc emulators like qemu, sheepshaver or pearpc.

[Crumb]

Still a great question and still not answered. Big endian. Small endian. Why not g5? Why not multicore support? And the bigger question to ALL amiga NG os devlelopers is this: Why not SMP? The goal of os 3.x api long now has been duplicated. How to write extenstions to that api to allow multi-core.

G5 is not ruled out but there are more important things that need all their attention. On MorphOS hardware is both available & cheap and reasonably powerful, all Amiga systems right now lack new apps, that's the main concern, not hardware support.

I would love a G5 version but my G4 Mac Mini runs very smoothly and so does my powerbook. If you haven't tried MorphOS yet you can do it right now if you want, you don't have to wait for any G5 version, it's very fast right now even on old G4s. If you find a G4 with nVidia graphic card you can reflash a Radeon more or less easily, there are tons of tutorials out there.

The problem for all AmigaOSes right now is lack of new powerful apps, not hardware support.

They all sleep on this very important question. Its SAD.

They are working on other important features like wireless support, R300 support, iBook support, improving the bundled apps, killing bugs... agreed that development goes slowly but I don't think there are many developers capable of doing that kind of job.

Yes hard to implement in current api. Someone must say that now so many machines have 4, 8, or even 16 cores. And all amiga os systems ignore this. Someone needs to write a new api.

Krashan explained that it wouldn't be difficult to add support for a new type of tasks focused on number crunching/memory access that could run on the other cores without causing any loss of compatibility, it would be some kind of AMP.