Desktop login hindered by broken software dependencies

I updated some software components once more on my Linux system today.
Unfortunately, I was surprised by the following information on my subsequent login retry for another graphical desktop session.

file:///usr/share/sddm/themes/breeze-openSUSE/Main.qml:169:26: Type Login unavailable
…
qrc:/qt/qml/org/kde/plasma/components/TextField.qml:12:1: Cannot load libary /usr/lib64/qt6/qml/org/kde/plasma/core/libcorebindingsplugin.so: /usr/lib64/qt6/qml/org/kde/plasma/core/libcorebindingsplugin.so: undefined symbol: _ZN11PlasmaQuick15SharedQmlEngine19setSourceFromModuleE14QAnyStringViewS1_
  • In which time range can the mentioned software dependencies become consistent again?
  • How will the chances evolve to switch to a working graphical user interface here?

Ask the developers at the appropriate places. This is the user forum.

Do you have additional 3rd party KDE repos enabled?

Can I achieve any related solutions for a questionable software situation?

  • Activation of an other Login interface
  • Clarification of contents in software packages

Will my bug report “Fix software dependencies for desktop login interface” become helpful accordingly?

Which packages are you holding back and why?

Mostly:

  • KDE Plasma version 5 component variants
  • Graphic chip support

:crystal_ball: long version transition

Which component variants?
What does “long version translation” mean?
How does graphics chip relate to particular Plasma components? Which GPU(s) do you have?

Likely I have more experience with zypper locks than anyone else here. Of my 20 or more TW installations that include Plasma, on only 3 have I allowed migration to version 6. My TW locks file is the same on each of those installations, currently @8,937 bytes. On none is SDDM installed. Given presence of sddm among your error messages, I suggest you give a different DM a try. Most of mine use either KDM3 or TDM, neither of which are in standard repos. 1 or 2 use LightDM. If switching DM doesn’t work, or even if it does, it may be that you can login on a vtty, then start a session using XDG_SESSION_TYPE=XXX dbus-run-session startplasma-XXX, where XXX would be wayland or x11.

Transition (for example):
KDE 3, 4, 5, 6 (Plasma) …

Such feedback sounds promising.

I can eventually try other display managers out to check if they provide a working login interface also on my system.
It depends on the aspect if corresponding parameters can be selected in a “convenient” way.

:thought_balloon: It seems that some contributors have got understanding difficulties with the reasons for temporarily locked packages.

Do what Mrmazda said, try another display manager. This is just speculation, not advice, but I wonder is holding sddm back could have prevented your issues.

:eyes: It seems that the selection of corresponding graphical user interface alternatives is working (as expected) also by the tool “update-alternatives”.

:thought_balloon: I became curious once more how involved software dependencies can be checked and improved further.

As already pointed out in the bugreport. There is nothing to improve regarding dependencies. They_are_working. You decided to break the dependencies by applying locks to packages without the appropriate knowledge of the consequences. As can easily be seen at your zypper ll list, you decided to lock random Plasma 5 packages. This lead to an inconsistent Plasma 5/6.0/6.1/6.2 system. So you shot yourself in the knee and try to blame the devs for it.

There is nothing what devs can do when some users decide to willingly break their system.

You were not even able to explain why you believe the locks for Plasma 5 are necessary. Instead of „answering“ questions from devs with unrelated counter questions, you could have easily checked within 5 minutes, that for all of your locked Plasma 5 packages the equivalent v6 packages are available since ages.

I am curious under which circumstances further clarification opportunities will get the desired attention for affected technical aspects.

:crystal_ball: Eventually?

I am influencing some software dependencies for a while.
:crystal_ball: Collateral evolution will happen accordingly.

There were “personal” lock selection criteria involved.

Partly, yes.

Perhaps.

No.

But I propose further considerations with the potential for desirable improvements.

Advanced tools can support safer and more convenient dependency management, can’t they? :thinking:

I tried to reduce temporarily unwanted questions for resolution options by a tool like “zypper”.

:thought_balloon: I got doubts for this view.

If you only accept answers you like, please don’t post at all. FYI, @hui’s statements are correct.

Do we eventually take different sets of software components into account for this discussion? :thinking:

This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.