This might indicate a missing signal being emitted when moving a ShellClient instead of a Client.
Looks wired up fine from (only location) AbstractClient::startMoveResize() I'm rather worried that painting happens in postPaintScreen, eg. I get superimposing lines on repaints...
> I'm rather worried that painting happens in postPaintScreen Yeah, I'm just changing all effects to the new traits api and while adjusting snaphelper I also noticed that and moved the code into paintScreen.
Git commit 42e82f09be74023d84361e77dd8ab289bef44b08 by Martin Gräßlin. Committed on 29/08/2016 at 05:53. Pushed by graesslin into branch 'master'. Move Q_PROPERTY definitions from Client to AbstractClient The properties: * maximizable * moveable * moveableAcrossScreens * resizeable Were only defined on Client instead of AbstractClient. This resulted in the EffectWindow having those properties evaluate always to false for a ShellClient and breaking some effects. M +24 -0 abstract_client.h M +8 -0 autotests/integration/shell_client_test.cpp M +0 -24 client.h http://commits.kde.org/kwin/42e82f09be74023d84361e77dd8ab289bef44b08