Jeremy Fitzhardinge wrote: > > I can deal with the change going into -git, but it does seem awkward > knowing that it is the wrong change and it will be replaced by something > else almost immediately. > Well, it is not quite wrong - it is appropriate for -git. That it will be replaced soon is a minor thing, as long as we can work together to make sure your patches are unaffected. > My main concern is that the Xen patch queue is complex enough as-is, and > I've been trying hard to remove dependencies on other uncommitted > patches. It's doubly complex because I'm not really sure if I'm > targeting Andrew or Andi's tree as a base, though at the moment -git > seems to work either way. > > J > Believe me, I understand the complexities of dealing with -mm, -git, and -i386 trees. It's not just doubly complex, its triply so. I believe the best approach is to make it easiest on the maintainers. So the patch stream I am uploading through -mm, -i386, to -git does not conflict with the Xen patch queue as long as we make sure the next bundle to those same trees applies clean. Again, I broke the Xen patch queue in -mm, and for that I am sorry. I will help you re-integrate against the VMI patches I sent out, as I broke your code, I will help fix it. And we both have a tree to work off from, which makes it rather easy to work together ;) Cheers, Zach ---- Sutra I.2 - interruptah chittavritti nirodhah
