https://bugzilla.redhat.com/show_bug.cgi?id=1335988 --- Comment #22 from Jerry James <loganjerry@xxxxxxxxx> --- (In reply to Antonio Trande from comment #21) > No; maybe is related to https://bugzilla.redhat.com/show_bug.cgi?id=1350911. That could be, but doesn't that imply that the symbols in libgslcblas are not used at all, just the symbols in the underlying blas library? Or maybe undefined symbols mess up the unused-direct-shlib-dependency check somehow. This looks much better. There is still an unowned directory problem, though. The giac-doc package puts files into the language-specific directories, which it does not own, namely /usr/share/giac/doc/{de,el,en,es,fr,zh}. Also, there are still a few issues from comment 17 that have not been addressed completely: - The spec file contains an ExclusiveArch tag with a note about PPC and aarch64. What are the issues? Is somebody working to resolve them? This information would be useful in a spec file comment. Also, is it only PPC and aarch64 that are not ready? If so, ExcludeArch should be used instead of ExclusiveArch. What about mips, for example? - Would it be possible/advisable to split out the noarch files under %{_datadir} into a noarch subpackage that is required by the main package? That would allow sharing across architectures. - There is still one hidden-file-or-dir warning. Is this file needed? giac-doc.noarch: W: hidden-file-or-dir /usr/share/giac/examples/Exemples/analyse/._signal.xws -- You are receiving this mail because: You are on the CC list for the bug. You are always notified about changes to this product and component _______________________________________________ package-review mailing list package-review@xxxxxxxxxxxxxxxxxxxxxxx https://lists.fedoraproject.org/admin/lists/package-review@xxxxxxxxxxxxxxxxxxxxxxx