/rant mode on
The big problem with OS3, OS4, MorphOS and AROS is (specially with the last one) lack of killer apps, not lack of fast hardware.
AROS for big endian machines (like PPC or ARM configured in bigendian mode) may be able to run 680x0 binaries transparently but most of AROS devs aren't interested in compatibility because they no longer use any amiga program (they use linux or windows).
The old 680x0 apps are better than nothing and many of them are still useful.
AfA is quite important because it helps to improve AROS because lots of people will use it everyday (and the bug reports can end up improving the AROS source tree)
An AROS-PPC release for PowerMacs with builtin 680x0 emulation and a MorphOS wrapper may be enough to keep people interested on running old apps happy (like me).
The biggest incoherency I find in most of AROS developers claims is that they aren't interested on running old apps. If that's true why are you using an obsolete design without memory protection, multiuser, etc? With AROS you have the worst of both worlds: you lack 680x0 compatibility but you also lack memory protection.
The only AROS advantage is that it is free and portable, but as people is more interested in reinventing the wheel (do we really need 4 different SDL flavours? yes... one done by Gabrielle Grecko for OS3.x 680x0, another one for AROS, another one for MorphOS, another one for OS4). And the AROS team instead of merging their code with the main tree of various MCC classes at sourceforge decide that it's better to merge the changes by hand each time an mcc class is updated (we are lucky Geit is trying to adapt SDI includes) and keep an AROS-only source in their repository.
Then they decide that the AmigaOS handler system sucks and implement their own incompatible system. I don't care if it's better or not, but if you think that it's better why don't you also implement memory protection and get rid of the old amigaos3.x api? at least most of AROS developers think that old amiga apps suck, so why keep a sucky API?
OS4 lacks hardware and due to the stupid Amiga Inc licenses will have no future unless the situation changes and hardware appears. MorphOS situation is better in the hardware side but it's clear that money is required to materialize MOS-team dreams about Q-Box (something I still don't understand is why the team implemented the drivers inside of the not-memory-protected A-Box instead of making a generic driver that wrapped its calls to the real drivers inside the Q-Box. At least the keyboard driver would have been nice so you could reset MOS when it hangs completely. That leads me to think that the q-box is so small and light that it almost doesn't exist)
AROS has some future, but without apps and with the current attitude of developers that give their backs to the current amiga community (that is also splitted also thanks to MOS/OS4/AROS coders who prefer to make own branches instead of adding their code to the main branch and having all versions created automatically) it will have even less future. Some open minded developers have ported their apps to AROS but I rarely see AROS developers porting themselves their apps to other oses (including os3).
Examples of community split:
-2 branches for Mplayer, one for OS4 and one for MOS when 98% of code is the same.
-4 different SDL versions (OS3, AROS, MOS, OS4)
-AROS devs not adding their changes to the sourceforge MCC classes that they use in Zune
-AROS devs not releasing native Zune versions for OS3 regarless of the fact that it started being developed on 68k
-Developers who think that aminet sucks and that everyone should check out their cool webpages who don't upload their stuff to aminet.
-4 or 5 pci libraries (AROS, Prometheus, Mediator, CyberPCI, CyberPCI for MOS, OS4 expansion library PCI interface... isn't great rewritting 4 times the same driver?)
-4 or 5 versions of expat.library, some of these released various times for the same OS (this is caused probably due to authors not using aminet...). I've seen 2 or 3 different expat versions for 680x0, 2 for OS4, another one for MorphOS, another one for AROS. Instead of making a single port and adding some ifdefs or using SDI headers we have various sources spreaded around the world...
-Ignorant AROS users who think that AfA is useless for AROS although it's the only way of testing it with real AmigaOS apps.
UAE for AROS is a worse compatibility solution than WinUAE for Windows. I mean, I prefer to run Windows+WinUAE than AROS-x86 and EUAE.
An aros big endian release for PPC or ARM (in big endian mode) would please some of us if it included 680x0 compatibility and also a MOS wrapper.
Anyway, we lack killer apps (and common apps like a decent browser. Marcik will deserve an statue when he finishes his KHTML port). That's the real problem.
For AROS/OS3/OS4/MOS most of hardware is fast enough. But all the hardware is useless without software. OS4 and MOS at least have 680x0 compatibility and can run useful apps.
Unless everyone that is writting apps for OS3/OS4/MOS turns their head to AROS and starts porting their software it will be a beautiful, fast, portable and efficient -useless- toy
When my amiga interest is low I concentrate on scene stuff (coding classic gfx effects, watching demos and intros...) and that keeps me happy.
Looking at AROS depresses me as much (or more) than OS4 or MOS
/rant mode off
The best thing an user or developer can do is simply enjoying what you have using it to the maximum instead of browsing forums or thinking about the future of the platform. It's just a computer, enjoy it now that it's working, before the capacitors acid eats all the motherboard wires...
