Thread: New Council elected! May 2012 - November 2012
View Single Post
Rob1n is offline Rob1n
2012-05-01 , 15:50
Posts: 3,617 | Thanked: 2,412 times | Joined on Nov 2009 @ Cambridge, UK
#302
Originally Posted by szopin View Post
Which is exactly why opensourcing maemo code could change a few things. Once we get ARM (with erratas) working, all future processors are a go (or is ARM going anywhere?). Sure drivers are the limiting factor, but once they are the ONLY limiting factor maemo on any hardware can be a reality. Things in mobile world do not change from atari to amiga to pc all the time. Tegras etc will be a problem, still paying for reveng of drivers would be much less costly than expecting whole system on each and every device (also upstream all reveng'es to linux kernel and more people will join this community, no?)
I think you need to look at the ARM tree within the kernel. Getting things working for one specific ARM processor will get not get you very far at all with the next one. Unlike the x86 world, there's very little standardisation - even when the ARM core is the same, the rest of the on-chip hardware can be radically different (or even be identical but use different addresses, etc. for the same components).
 

The Following 5 Users Say Thank You to Rob1n For This Useful Post: