Yep, that's what many of us have been trying to figure out. Having emulation layers to support non-native software in place of native software isn't better, it isn't new and it certainly won't fly in the serious market. So it looks like: a) it'll just be Linux with a more thorough skinning than the one on the C64x, maybe a new desktop at a push, b) a totally new OS with emulation to replace the lack of native software and modern features, which will be a marketing disaster, or c) CUSA have the funds to bankroll dozens of developers for several years to come up with a genuinely viable OS and ecosystem, and the marketing, support and distribution of it all.
Place your bets now please.
I wasn't aiming at that... I was more speaking of Linux, more specifically Linux kernel and why would one need to write a new one.
If you go any other route than Linux, then you're bound to end up in a porting hell just like AOS, Syllable, Haiku and similar OSes are doing.
Very few native developments, mostly software ported from Linux/Windows.
In that situation, one musk oneself. Do I really need to invest and make a completely new OS just to run mostly the same apps/games that I'd do on Linux, but more buggy and slower? Obvious answer is obvious.
Naturally, if you're doing it for fun and hobby, why not. I'd do it. If you wanna be serious, UNIX/Linux is what is needed today.
I can't think of any major, successful OS(including desktop and mobile) in recent years that wasn't based on UNIX/Linux or wasn't POSIX, besides Windows. Please correct me if I'm wrong.