As amarok is not currently shipped for x86_64, kadu-amarok is useless for it too. Attached is a quick patch that disables building it on archs that amarok is not currently buildable for - list of archs grabbed from the amarok package.
Created attachment 137693 [details] Don't build -amarok on archs that have no amarok
The useless kadu-amarok subpackage has been removed from the devel package repository (not CVS). Please make sure to disable it before the next build.
Forgot to note that the subpackage has been removed only from the x86_64 repository, it's still in the others.
Fixed in actual release. Closing it.
Even though the fix applied in 0.5.0-0.14 seems to work, it looks very strange to me - I can't say offhand why it works in the first place, and it doesn't actually disable building the amarok stuff, it just %ghosts the files which means they're probably still included in some filelists. The suggested fix in comment 1 would have been cleaner and easier to understand at least to me... Anyway that's not the real reason why I'm reopening this bug. The bigger one is that this needs fixing in FC5 too now, amarok got the excludearchs there too.
(In reply to comment #5) > Even though the fix applied in 0.5.0-0.14 seems to work, it looks very strange > to me - I can't say offhand why it works in the first place, and it doesn't > actually disable building the amarok stuff, it just %ghosts the files which > means they're probably still included in some filelists. The suggested fix in > comment 1 would have been cleaner and easier to understand at least to me... Before I'll make that fix to FC5 branch I have to explain why have I used %ghost macro. Well, comment 1 fix caused a build system error: RPM build errors: Installed (but unpackaged) file(s) found: /usr/lib64/kadu/modules/amarok.so /usr/share/kadu/modules/amarok.desc /usr/share/kadu/modules/data/amarok/amarok.png /usr/share/kadu/modules/translations/amarok_de.qm /usr/share/kadu/modules/translations/amarok_pl.qm so I had to find the best way to make it working. Using of %ghost macro works well and doesn't cause any rpmbuild error. As you see, I tried to fix these errors a lot of times (just have a look on changelog) and that way looks the best for me. If you have any other ideas, feel free to write it here.
(In reply to comment #6) > Before I'll make that fix to FC5 branch I have to explain why have I used > %ghost macro. Well, comment 1 fix caused a build system error: Note that you seem to have tried the patch in comment 1 only partially, compare these: Submitted one: https://bugzilla.redhat.com/bugzilla/attachment.cgi?id=137693 Tried one: http://cvs.fedora.redhat.com/viewcvs/devel/kadu/kadu.spec?root=extras&r1=1.6&r2=1.7&makepatch=1&diff_format=h The change you tried does not contain this change: +%ifnarch ppc64 x86_64 s390 s390x ia64 module_amarok \ +%endif I test-compiled the patch in comment 1 on x86_64 before submitting, it worked here as expected.
You're obviously right - that was my fault. Now it's fixed in your way.