Since there are no objections, let's move ahead with this. On Fri, Mar 24, 2023 at 7:36=E2=80=AFPM Nate Graham wrote: > > Hello Ivan, > > I think this may be a more pertinent question for Plasma folks; CCing > plasma-devel. > > Nate > > > > On 3/22/23 10:41, ivan tkachenko wrote: > > Hi, > > > > I'm working on Flatpak KCM. > > > > This is all very broken and we can't do conservative fixes because they > > won't > > apply well between master and Plasma/5.27 branches if done > > little-by-little, > > and because the thing is fundamentally broken on multiple levels. I man= aged > > to refactor and rename members, methods and model roles, so they at lea= st > > make sense and don't duplicate each other too much. I also started writ= ing > > tests and expanding test suite on newly found broken cases and introduc= ed > > API. But this is not enough, and on my roadmap there are things like: > > > > - Replacing all section headings from partially(!) translated strings > > strings+enum heuristic into a single (and exposed to QML) enum; > > - Rewriting value lists for ComboBox'es from pre-translated QStringList > > into > > model with text and value roles; > > - Rewriting all model setters (both toggling and changing string values= ) > > from > > storing override data in manually formatted ini/KConfig style string > > into > > storing proper data structures and value identifiers inside individu= al > > entries and serializing them in a centralized manner; > > - Reading "defaults" not just from metadata, but from other three overr= ide > > levels not writable by KCM (system-default, system per-app, > > user-default); > > - Fixing model signals, like using beginInsertRows/endInsertRows instea= d of > > just dataChanged approriately; > > - Changing UI styling for entries that are marked for removal, such as > > unchecked non-default filesystems, D-Bus services & prefixes, and > > environment variables > > - Implementing default settings highlighter. > > > > Can I please have permission to do aggressive cherry picking from maste= r to > > make it better? > >