3 ms·
The success rate for new ground-up modern OS projects achieving broad adoption isn't great - Microsoft's Midori, GNU Hurd, Apple's "Pink", etc. AFAIK there bas
by contextfree 10y ago
The success rate for new ground-up modern OS projects achieving broad adoption isn't great - Microsoft's Midori, GNU Hurd, Apple's "Pink", etc.
AFAIK there basically hasn't been a broadly successful one since Windows NT in 1993. Which isn't to say people should stop trying and it'll be fantastic if they do succeed with this, I just wouldn't count on it.
- harigov 10y agoIt may not be so bad because most of Android apps are built on Java which isolates a lot of what OS does. For other native apps, if Google builds a POSIX compliant OS, it should be fairly easy to port/run.
- mathattack 10y agoI recall reading a book about the development of NT. One of the key figures joined in because it was seen as the last great chance to create an OS from scratch. Seemed very short sighted to me at the time.
- coldtea 10y agoThat's because those weren't completed in time due to feature creep and second system effects (Apple's Pink/Copland), lack of interest and slow progress (Hurd), etc. Midori is just a research project, so noone pushed for it to be adopted anyway. With a completed OS (which nowadays isn't so difficult, large parts of the stack are either ready or there for the taking), and a big vendor behind it, it's another story. Microsoft pulled it off with NT, and Apple's OS X was a whole new OS as far as Mac customers were concerned (the fact that it already existed as NeXT didn't matter much -- it still required all new software (it didn't run old NeXT stuff as is), new drivers, had a new SDK that Mac developers had to cater too, etc. Besides Android is not like a full desktop OS Linux -- it abstracts most things for apps through its SDK/"Java", and Linux is just the kernel.