Hello Bernhard, yes, thanks, it probably is the same, although in my journal I have traps from kded5 and not from kde5, but this is probably a copy-paste issue in #996726. Unfortunately, it's not mentioned, what the fix was, but when I look at the diffoscope of 4:5.23.0-2 and 4:5.23.0-3 it looks like 4:5.23.0-3 would solve this issue, too, as the main change is to tighten the breaks >Breaks: kwin-common (<< 4:5.22.90), kwin-decoration-oxygen (<< 4:5.22.90), kwin-style-breeze (<< 4:5.22.90), kwin-wayland (<< 4:5.22.90), kwin-wayland- backend-drm (<< 4:5.22.90), kwin-wayland-backend-fbdev (<< 4:5.22.90), kwin- wayland-backend-virtual (<< 4:5.22.90), kwin-wayland-backend-wayland (<< 4:5.22.90), kwin-wayland-backend-x11 (<< 4:5.22.90), kwin-x11 (<< 4:5.22.90) So the specific problem is solved. The question is, whether the implicitly mentioned general problem should be solved, too. Even with this change you'd get a mix of 5.21 and 5.23 when installing a specific kde-plasma-desktop package (see my list above in 996744). In this case this mix seems to work, however, do we really have the man-power to support these mixes in the future? And do we think that's what the user wants? Even in #996726 it's mentioned, that you should always upgrade the entire stack. Why don't we set the package metadata to guarantee that for the user and use our time for fixing real problems instead of package version mismatches? Kind regards Patrick
Attachment:
signature.asc
Description: This is a digitally signed message part.