https://bugzilla.redhat.com/show_bug.cgi?id=1460917
--- Comment #5 from Jakub Kadlčík jkadlcik@redhat.com ---
What about the name dist-git-client?
Personally I have no problem with both rpkg-client and dist-git-client. May clime use what he likes best (even some other name)
Source0: rpkg-client-git-4.88886a6.tar.gz
It is not docummented, how can we obtain this archive. In the first version of your spec file there was a
# Source is created by: # git clone https://pagure.io/rpkg-client.git # cd rpkg-client # tito build --tgz
if it is still valid, can you please put it back?
%package -n rpkg
This will create a subpackage called 'rpkg' (not 'rpkg-client-rpkg') which I believe is still considered as name conflict. Can anyone confirm or deny that, please?
Also with
%files -n rpkg
and no %files section for the whole package means, that the rpkg-client RPM is not even built.
I am sorry, I don't understand the whole rpkg subpackge idea, can you please clarify it for me? As I see it, the purpose of this package is to ship a single executable python script `%{_bindir}/rpkg` which imports `pyrpkg` and therefore it requires python2-rpkg. That's basically it. Why to use subpackages at all? Can't we have just a simple rpkg-client package which has the (Build)Requires and %files itself?
I see that you want to stick to generating a RPM named 'rpkg' for some reason, but I don't understand why. Why is it better than RPM named 'rpkg-client'?