| Summary: | kde4/plugins/styles/oxygen.so is not multilib | ||
|---|---|---|---|
| Product: | [Fedora] Fedora | Reporter: | Julian Sikorski <belegdol> |
| Component: | kdebase-workspace | Assignee: | Than Ngo <than> |
| Status: | CLOSED ERRATA | QA Contact: | Fedora Extras Quality Assurance <extras-qa> |
| Severity: | unspecified | Docs Contact: | |
| Priority: | unspecified | ||
| Version: | 15 | CC: | fedora, jreznik, kevin, ltinkl, rdieter, rnovacek, ry, smparrish, than |
| Target Milestone: | --- | ||
| Target Release: | --- | ||
| Hardware: | Unspecified | ||
| OS: | Unspecified | ||
| Whiteboard: | |||
| Fixed In Version: | Doc Type: | Bug Fix | |
| Doc Text: | Story Points: | --- | |
| Clone Of: | Environment: | ||
| Last Closed: | 2011-07-20 16:44:45 UTC | Type: | --- |
| Regression: | --- | Mount Type: | --- |
| Documentation: | --- | CRM: | |
| Verified Versions: | Category: | --- | |
| oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | |
| Cloudforms Team: | --- | Target Upstream Version: | |
|
Description
Julian Sikorski
2011-05-15 13:00:33 UTC
To be clear, as discussed on IRC, the main kdebase-workspace actually is multilib at this time, however it probably shouldn't be and it's full of multilib conflicts. I'll take a look. Changelog: * Sun May 15 2011 Rex Dietr <rdieter> 4.6.3-5 - kde4/plugins/styles/oxygen.so is not multilib (#704840) I'm afraid testing will need to wait until KDE 4.6.3 is in one of the repos; trying to fetch all the packages from koji seems like a neverending dep hell ;). It'll be in kde-testing from http://kde-redhat.sourceforge.net/ shortly. Will be submitting 4.6.3 to updates-testing official real soon now. > Changelog: > * Sun May 15 2011 Rex Dietr <rdieter> 4.6.3-5 > - kde4/plugins/styles/oxygen.so is not multilib (#704840) I don't think this is going to fix the whole problem. There's still the bug that kdebase-workspace.i686 (as opposed to -libs) ends up in the x86_64 repo even though it's full of multilib conflicts. Or is that because of the Oxygen style being whitelisted somewhere? I think at least these files: /usr/lib/libkickoff.so /usr/lib/libpowerdevilcore.so /usr/lib/libpowerdevilui.so /usr/lib/libsystemsettingsview.so belong to -libs. There are also some /usr/lib/libkdeinit4_*.so files, I'm not sure what the multilib heuristic does on those. (They should really be blacklisted for multilib.) > Will be submitting 4.6.3 to updates-testing official real soon now. FYI, you can submit 4.6.3 to F15 testing now, I don't think we have any remaining blocker or NTH bugs open against 4.6.2. I can confirm that kdebase-workspace-4.6.3-5.fc15.x86_64 fixes the problem, but only partially. I still need to run $ export QT_PLUGIN_PATH=$QT_PLUGIN_PATH:/usr/lib/kde4/plugins/ Would it be feasible to add this to plugin path automatically? We used to patch qt to unconditionally do that, but it caused some bad side-effects. I've an idea for something a little safer. 1. patch qt to append /usr/lib/qt4/plugins if not already present 2. patch kde (kdebase-workspace/startkde) to append /usr/lib/kde4/plugins if not already present As discussed on irc, 1. makes little/no sense, this is all about QT_PLUGIN_PATH (in startkde) Though, I had hoped to be able to handle the Qt outside of KDE case, but that's more of a corner-case. %changelog * Sat May 21 2011 Rex Dieter <rdieter> 4.6.3-7 - multilib QT_PLUGIN_PATH (#704840) F15 build: http://koji.fedoraproject.org/koji/taskinfo?taskID=3085590 This indeed helps, thanks. |