Re: Highlighting totem-upnp issue and possible koji problem?

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Thu, 2011-05-26 at 17:48 -0300, Michael Wiktowy wrote:

> I can understand the whole build not being killed over an optional
> plugin. However, should koji build an empty package? I would think

It rather depends on the spec file.

If you look at the spec, it has this:

%files upnp
%defattr(-, root, root,-)
#%{_libdir}/totem/plugins/coherence_upnp

so clearly the packager noticed this problem, because the file list for
upnp is now effectively empty, with that line commented out. So rpm and
koji have only done what the packager told them to. Thanks to git blame
(God, I love that tool) we can see that Bastien made this change on
2010-11-02, but the changelog just says "Fix up filelist", which is a
bit cryptic. So I don't see any tool fail here, just a change by the
maintainer which perhaps wasn't prominently enough described.

In case you don't know, rpm is pretty strict about package file lists,
in part so that maintainers don't miss this kind of thing happening:
exactly every file produced by a build has to be listed - i.e.
explicitly made a part of one of the packages produced by the build. The
build will fail if there is a file produced by the build which isn't
explicitly made part of a package, or if there is a file listed in one
of the packages which *isn't* produced by the build. So when the upnp
plugin stopped working, the package build would have started failing and
complaining that there was 'no such file
%{_libdir}/totem/plugins/coherence_upnp' , and until Bastien commented
out that line in the spec, the build wouldn't have worked. So there are
sufficient tool safeguards here to make sure that the packager must
_know_ the package is empty where it wasn't before, and accept that. 

You could say Bastien should have stopped the upnp package building at
all rather than making it empty, but then if he didn't make something
obsolete it, old versions would have stuck around and caused trouble,
and if he makes something obsolete it, then re-introducing it when it
actually works becomes more complex.
-- 
Adam Williamson
Fedora QA Community Monkey
IRC: adamw | Fedora Talk: adamwill AT fedoraproject DOT org
http://www.happyassassin.net

-- 
devel mailing list
devel@xxxxxxxxxxxxxxxxxxxxxxx
https://admin.fedoraproject.org/mailman/listinfo/devel


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[Index of Archives]     [Fedora Announce]     [Fedora Kernel]     [Fedora Testing]     [Fedora Formulas]     [Fedora PHP Devel]     [Kernel Development]     [Fedora Legacy]     [Fedora Maintainers]     [Fedora Desktop]     [PAM]     [Red Hat Development]     [Gimp]     [Yosemite News]
  Powered by Linux