DevHeads.net

F30 live images still contain dbus-daemon package, anaconda-core requires it ?

Hi All,

I just noticed that Fedora-Workstation-Live-x86_64-30-20190408.n.0.iso still
contains the dbus-daemon package even though we are using dbus-broker now.

I've tried to rpm -e it and it for some reason it is required by
anaconda-core which seems weird.

So should I file a bug against anaconda for this? I guess it is too late
to fix this for F30?

Regards,

Hans

Comments

Re: F30 live images still contain dbus-daemon package, anaconda-

By Martin Kolman at 04/10/2019 - 13:12

On Wed, 2019-04-10 at 11:07 +0200, Hans de Goede wrote:
Adding Jirka Konecny who handled this on our side to CC.

Re: F30 live images still contain dbus-daemon package, anaconda-

By =?UTF-8?Q?Ji=C5... at 04/11/2019 - 03:39

On Wed, 2019-04-10 at 19:12 +0200, Martin Kolman wrote:
Hi, yes it is exactly what Martin said.

I've spend a lot of time to find a way how we can migrate to DBus-
broker but the result wasn't something worth doing now.

The problem is that Anaconda have to start it's own session of dbus but
dbus-broker, by design, don't have easy way how to do this without
systemd running. And unfortunately, Anaconda is also used by RCM in the
environment (old chroot mock), where is no systemd and it is not that
easy to start systemd there.

The result would be to replace existing launcher and play with local
sockets for basically no benefit and broke compatibility with RHEL. The
broken compatibility would make our life harder if we decide to make a
rebase. We can do that later but for now we have a plenty of other
work.

Hope the explanation helps you,
Jirka

Re: F30 live images still contain dbus-daemon package, anaconda-

By Michael Catanzaro at 04/11/2019 - 11:21

On Thu, Apr 11, 2019 at 2:39 AM, <a href="mailto: ... at redhat dot com"> ... at redhat dot com</a> wrote:
TBH we should probably keep dbus-daemon around until we have a
replacement for dbus-run-session anyway; it's too useful when working
with headless environments like continuous integration, containers,
etc. There is already an issue for this at
<a href="https://github.com/bus1/dbus-broker/issues/145" title="https://github.com/bus1/dbus-broker/issues/145">https://github.com/bus1/dbus-broker/issues/145</a>, so Anaconda is far from
the only one having this problem.

Michael

Re: F30 live images still contain dbus-daemon package, anaconda-

By Chris Murphy at 04/10/2019 - 13:03

On Wed, Apr 10, 2019 at 3:07 AM Hans de Goede < ... at redhat dot com> wrote:
Anaconda. Freeze is next week.

Re: F30 live images still contain dbus-daemon package, anaconda-

By Adam Williamson at 04/10/2019 - 13:14

On Wed, 2019-04-10 at 11:03 -0600, Chris Murphy wrote:
It's not a bug. All is as intended.

<a href="https://bugzilla.redhat.com/show_bug.cgi?id=1658739#c9" title="https://bugzilla.redhat.com/show_bug.cgi?id=1658739#c9">https://bugzilla.redhat.com/show_bug.cgi?id=1658739#c9</a>