Alternatively, don't toLower when wriing the rule instead of lowercasing (utf-8) on each and every rule check?

In D2574#48138, @luebking wrote:

Alternatively, don't toLower when wriing the rule instead of lowercasing (utf-8) on each and every rule check?

I cannot find a toLower when writing the rule - neither in rules.cpp nor in kcmkwin/kwinrules

kcmkwin/kwinrules/main.cpp:86
QByteArray role = info.windowRole().toLower();

The value is first used to compare against existing rules and afterwards (in case) to build a new one.

In D2574#48142, @luebking wrote:

kcmkwin/kwinrules/main.cpp:86
QByteArray role = info.windowRole().toLower();

The value is first used to compare against existing rules and afterwards (in case) to build a new one.

ahhh. I only looked at the kcm code where that's not the case. So when going through systemsettings it's not going to be lowered, but when going through useractions menu, it's going to get lowered.

Interesting. So yes that would be a solution. Though it doesn't "fix" existing rules.

No, and they're not post-fixable at all (restore case)
Not sure why any case insensivity was added itfp, though (makes sense to ease hand-written rules, but that's not a regular task at all)
*shrug*

I can't really comment on this solution as I can't judge the regression risk, I'd say shipit if you're comfortable with it, or abandon if not.