Did you pass bugs and paches to the orginal projects?

Dan Nicholson dbn.lists at gmail.com
Sat Jan 27 11:28:30 PST 2007


On 1/27/07, Nadav Vinik <nadavvin at gmail.com> wrote:
>
> The LFS book use in paches in order to build full system since the
> orginal parts of linux system apparently don't realy work together.

It's not really documented, but we try to when the patches are
applicable upstream. Some patches are specific to building LFS, like
the GCC specs patch in Chapter 5. We're building GCC in a special way
that's not relevant to the GCC developers.

Some of the patches are from the original projects. They're just not
released yet, or from newer releases than we're using. We call these
backports. This includes something like the module-init-tools patch.

Some of the patches are specific to the way that we're trying to build
LFS. Submitting them back to the developers might not be worthwhile
since it's not the general consensus that the patch is the right way
to go. A lot of the UTF-8 fixes fall into this category. Sometimes
they've been actually rejected by the developers.

Sometimes, the upstream project is dead, and applying certain patches
is just an accepted part of building a package since there's no place
for the patches to get submitted to. I don't think this happens much
in the LFS projects since if an upstream project is dead, we usually
try to find a new active project.

We try to submit the useful patches back upstream. Sometimes they're
ignored or rejected, though. This is no different than any other
distribution. It can actually be very hard to take all the suggestions
from different parties and find a common fix that works for everyone.

Hope that explains things a bit.

--
Dan



More information about the lfs-dev mailing list