With ChromeOS, Google will have stronger incentives to work with the kernel community than what they have with Android.
The reason is that while Android is available only for very specific hardware, ChromeOS has the potential to be available for any PC. People are unlikely to upgrade their phone with a new network card (or whatever) and then go complaining to Google when it doesn't work, but they *could* do that with their ChromeOS devices.
For the ChromeOS Linux kernel Google has two choices:
1. Working with the community and having the community doing most / all the porting work for free.
2. Working without the community and as time goes by having to put more and more resources into keeping up with the upstream kernel.
Working with the community sounds like the cheaper option here in the long run. If Google has long term plans for ChromeOS, I can't see what they would gain by not letting the community work for them.
Posted Jul 16, 2009 10:39 UTC (Thu) by kragil (subscriber, #34373)
[Link]
_*I*_think_ Google will be much more upstream friendly with this one. I can imagine that they will go rolling release and actively push new kernels etc. They could use a snapshotting FS for desaster recovery. I think they will use an OpenGL WM and they will use KMS, PulseAudio and maybe even Ksplice because they want to push updates silently without user notification. I should just work and not bother people with technicalities.
I think such an approach is sufficiently different that it might work.
( Oh and if they do as I predict they will probably also use an established LSM framework for security, but http://code.google.com/p/nativeclient/ for native code )
But I suck at predictions .. my prediction for 2009 were e-ink ebook readers everywhere and books dying and other stuff that might still take an century to make a dent.