What does package 'maintenance' mean?

[David Woodhouse]

On Sat, 2007-05-05 at 18:09 -0600, Jonathan Steffan wrote:
> David Woodhouse wrote:
> > So? Surely it's the rôle of the Fedora package maintainer to _make_ it
> > work with Fedora? Is the package maintainer AWOL?
> 
> I'm not AWOL and I don't think it is my job to fix upstream code.
> Ok, I will state yet again; This is not a packaging issue. 

I appreciate, having looked at the references you provided, that fixing
this would require a 'fair bit of work' and that you're a volunteer;
it's not unreasonable that you haven't done it all by yourself within
the last year -- I'm not suggesting otherwise.

But let's address the _general_ case (and I've changed the subject
accordingly to avoid that specific topic behind).

I believe very strongly that it _is_ the package maintainer's job to
work with upstream code to make it work with Fedora, and this kind of
thing _is_ a packaging issue.

There's a reason we have Fedora package maintainers instead of just
automatically pulling in upstream tarballs and building them with
rpmbuild -ta. It's because the rôle of the package maintainer is to make
the package a _part_ of Fedora -- that's what makes Fedora a coherent
distribution and not just a semi-random collection of packages.

That means both local modifications and working with upstream where
necessary -- not just to ensure that the package is configured properly
for Fedora, but also that it _works_ properly as a Fedora package --
that IPv6 support works, that it works with exec-shield and SElinux,
that it builds and works on all supported architectures, and most
importantly that it builds and works with the other versions of software
which are included in Fedora.

In many of the examples there, there are resources available to assist
package maintainers -- people with experience of particular languages or
architectures, SElinux, IPv6, etc.

But it _is_ the job of the package maintainer to ensure that the package
actually works as part of Fedora -- asking for help where it's needed,
by all means, but taking charge of the process and working with upstream
to make it happen.

I really don't like to think what would happen if our kernel, gcc, glibc
and other package maintainers all suddenly declared that their job was
just to package what's upstream and not guide its development.

-- 
dwmw2