On Sun, 3 Apr 2011 16:16:21 +0100 David Goodenough <david.goodenough@btconnect.com> wrote: > The big difficulty with this is going to be getting all the fixes that > are necessary into the kernels. Given that normal Debian kernels would > not need the fixes what is their incentive to accept them. Just look at > the enormous list of patches that OpenWRT applies to the kernels, many > of which are for SoC peripherals (and this applies to MIPS SoCs as well > as ARM ones). Yes these fixes should be in the kernel, but they are not > and without them the kernels are no use for boards like Micro Routerboards and > Ubiquity Routerstations (to take but two I have tried). It's wider than that. We don't need to consider getting these fixes into Debian patches to the kernel IMHO. We need to consider how to aggregate those patches across a variety of boards and get that patch set into the mainline. This is wider than just Debian and Emdebian - this reaches across the entire embedded community. The current patches look like submissions from a bunch of individuals, not a community at all. To me, this sounds like a job for an umbrella organisation - maybe Linaro? -- Neil Williams ============= http://www.linux.codehelp.co.uk/
Attachment:
pgp4rYkCOH6zX.pgp
Description: PGP signature