ryan.oliver at pha.com.au
Sun Nov 27 20:00:01 PST 2005
On Sun, 2005-11-27 at 19:18 -0800, Jim Gifford wrote:
> Jeremy Huntwork wrote:
> > Jim, I understand your desires and goals, but this is not a Cross-LFS
> > issue. It's really an LFS issue, and it should be worked on there. If
> > anything, it should be done in a new branch of LFS trunk, not here.
> > Please don't bring this type of change only into cross-lfs simply
> > because you can.
> This affects Cross-LFS in a big way, udev rules and the hotplug we
> currently use, does not address all the hardware for all our platforms,
> it also will fix a known hotplug issue with MIPS and ARM based machines,
> that's why this is being investigated and going to be fixed on way or
> another. That's the main reason for Cross-LFS implementing this as soon
> as possible. Every one who has tried to build a modular system on MIPS
> will understand, the problems.
My question here is, what is the problem in doing development in an
"unstable" tree which does not yet have a release?
We need to fix issues as we hit them, and with the architectures we are
covering we tend to run into a few more quirks than standard lfs does.
I personally cannot take issue with forward planning to resolve issues,
especially when done in conjunction with upstream maintainers, my view
is it should be encouraged.
As to where this work should be done (LFS or Cross-LFS), does it really
matter? Should cross-lfs wait for lfs, ignoring broken behaviour on
certain architectures, or actively go about working towards a solution?
More information about the cross-lfs