On Fri, 2018-08-10 at 13:16 +0200, Erik Skultety wrote:
To be honest, I kinda was counting on you sending a follow-up
regardless of
my comment, because as you said, it's always better to document this properly
in relevant files :). At the same time though, I needed to experience e.g. the
failure with missing xz-static on CentOS 7 myself to actually figure out, what
it is that makes the build fail. So, even though I'm not insisting on you
changing the commit messages, I think that dropping a note about specific
missing requirements for a successful build as a justification for those 2
patches in play won't add much work. Anyhow, my RB still stands.
Note that the ultimate source of truth for which projects can be
successfully built on which platform is the Jenkins configuration,
not the guest configuration, and I'm not touching the former at
all because we were (correctly) already excluding platforms where
we knew the build would fail; this commit and the following one
merely clean up an oversight which caused a few unnecessary
packages to be installed in guests.
--
Andrea Bolognani / Red Hat / Virtualization