@deadwood
yes that is how I see it too
Aros became pure "research" but not really "replacement"
Users do not care about advanced USB support as long the whole system is not stable and there are bugs that make normal use impossible. The system must run stable and (as far as possible) bug free, it must be easy to install and hardware must be supported. As long this is not the case most users will drop it immediately
My idea (being no OS developer of course) concentrate on hosted versions regarding X86/X64 (Windows, Mac, Linux) with full support of the underlying OS (sound, USB, 3D, Network) and on V4 (68k) and RPi (ARM) native (because hardware is relative static)
Absolutely - as you say, relatively static hardware gives the best chance of success. What about this as a scenario?
- For those of us that like to run native, support one modern, barebones, relaively cheep and easily available x64 motherboard with plenty of PCIx express slots to add supported sound, network and graphics cards.
- For people who just want to use their current x64 machine, the options would be VM or hosted under Linux.
- For portable AROS, RPi 4 with AROS ARM native and one of the many kits for making it into a tablet or laptop.
- If possible (I appreciate USB is a problem here), AROS native on Pi 0 for embedded applications, hobby electronics and tinkering.*
- For the retro/gamer crowd, AROS 68k on Vampire.
Of course, the 'holy grail' for this is that all run the same API, so (within hardware limitations) one source-code file can be compiled for any of these targets Of course, as to how technically possible this is I am totally unqualified to say; it's just my musings.
Cheers,
Nigel.
*I still think there's a big opportunity for AROS to shine in this area. I'm currently setting up our annual Halloween haunted house full of animatronics triggered by various sensors, all controlled by AROS via Hollywood.