https://bugzilla.redhat.com/show_bug.cgi?id=1507103
--- Comment #84 from Fabio Massimo Di Nitto <fdini...@redhat.com> --- (In reply to Jan Pokorný from comment #83) > There's a misunderstanding, "%files -n libknet1-devel" comment should > stay where it was in 1.1.4. > > I was asking for a new one to explain the interim character of extra > treatment of debug packages that shouldn't have been introduced in > Fedora context in the first place. this is already addressed in comment #80 > > * * * > > re [comment 77], I am not familiar with how the test suite is run > for kronosnet, an example command would be "make check". > Nice-to-have category, though, the comment already explains why it > is not so straightforward in this case to run the tests. executing the test is straight forward make check, but we comment it out for safety. > > * * * > > Thanks for dealing with lz4 issues. > > Regarding "pkgconfig(openssl)" expression of dependencies, yes, they can > be versioned as well and/or can be combined with "Suggests" to prioritize > particular underlying package name should the conflict on such virtual > provides arise: > > https://fedoraproject.org/wiki/Packaging:WeakDependencies#Real_life_example > > Depending on how compat packages are structured, the same "satisfied by > more packages" situation could occur also with the previous cryptical > select-by-header-file approach, so there's effectively no regression > in this comparison. We will just switch back to BuildRequires: package-name. In context, upstream is also moving away from file based dependencies. -- 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@lists.fedoraproject.org To unsubscribe send an email to package-review-le...@lists.fedoraproject.org
