davidedmundson added a comment.
> So you oppose preloading of KRunner service, but accept it continuing to run forever awaiting its next use. I don't find this consistent. It's a common lazy-load pattern. Though that extension of closing down would also be a direction I would happily support. > This can't work fast. Any profiling to support that? I spent a few seconds and can see that we're still using PC2, we're loading not two results views which could happen later. One of the plasma framesvgs is emitted it's changed on startup, there's a blocking gap where nothing is happening... If this approach was suggested after extensive hours of profiling and optimisations, I'd have a different view, but it shouldn't be our first step. REPOSITORY R120 Plasma Workspace REVISION DETAIL https://phabricator.kde.org/D26718 To: pdabrowski, #plasma, #plasma_workspaces, davidedmundson, ngraham, broulik Cc: plasma-devel, Orage, LeGast00n, The-Feren-OS-Dev, jraleigh, zachus, fbampaloukas, GB_2, ragreen, ZrenBot, ngraham, alexeymin, himcesjf, lesliezhai, ali-mohamed, jensreuterberg, abetts, sebas, apol, ahiemstra, mart
