> I was very gently nudged...told that "I should stick w/ 2.4.21-behnxxxxx, as
> that's cutting edge for powerpc." I have to admit that I'm rather indignant
> about this "advice." (on this list, no less!) However, I have been told via
> irc that woody doesn't have a sufficient tool chain to properly compile
> 2.6.0-testx kernel yet, and elsewhere that the linux kernel is now an
> exclusive community of only developers, and that us peons who want vanilla
> kernels to compile for their (non-i386) platforms just have to wait.
Not having tried to compile any recent Linus kernel on powerpc or
elsewhere, I might be wrong, but from my experience it's always been the
case that different architectures each had their own development trees and
their own (rather small) development communities. It's always been up to
you to figure out where to get cutting edge kernel sources, and to deal
with problems yourself as they show up. Advice like 'better stick with
2.4.21-benh' might just mean that - unless you want to spend your time
fixing patch rejects or compile errors and risk kernel crashes, just use
the stable powerpc branch (which happens to be maintained by Ben).
Otherwise, join the exclusive community of kernel developers (they keep
secret resources such as the linux-kernel and linuxppc-dev mailing lists)
and join in to the fun. More kernel hackers working on less popular
architectures like the powerpc means faster support and faster integration
of these architectures in the mainstream source. If BenH has no time to
work on 2.5/2.6 and you absolutely need 2.6, better move fast. If you're
not skilled in kernel hacking, why bother with 2.5 or 2.6?
Another secret piece of information: Linus' tree has always been pretty
useless on the less popular architectures. Use Alan Cox' branch as a
starting point if there's no proper powerpc devel tree (which I doubt).
About the only gripe I have with current kernel development is the switch
to bitkeeper. But I've not looked very hard for current rsync access
either. I'm sure BenH and Paul do a great job in keeping up with new
hardware and kernel versions. I know how hard it is to keep current with
only a handful of kernel hackers even when hardware isn't a moving target.
Not sure what the toolchain stuff is about, do you have any details on that?
Michael
