I was previously told on this list
http://lists.qt-project.org/pipermail/development/2016-April/025595.html
that this bug is a KDE issue, not Qt. So I moved it to KDE list.
A dev @ KDE then said it is NOT a KDE issue, but IS a Qt bug
This is https://bugreports.qt.io/browse/QTBUG-48795
Qt5 builds it's internal events as combination of (synthetic) keys and HW
modifiers.
That is *clearly* a fucking bug in Qt: either it ignores synthetic
events (a regression compared to Qt4) or not, but just creating random
events out of different sources should certainly not be intended.
Another KDE dev agreed, and opened this issue
https://codereview.qt-project.org/#/c/156451/
Since then, after some daily pkg updates, I saw an improvement in the behavior
of shortcuts.
I'm not going to chase the bouncing "it's not us" ball anymore. in case
anyone's interested, I'm posting what I see here.
Here's a test showing that currently using shortcuts is still not 100%.
Trigger: Meta-Ctl-Alt-T
Action Expected Output Actual Output Status
--------------- ----------------- --------------- --------
1 1 1 OK
Shift+1 ! ! OK
1:1 11 11 OK
1:Shift+1 1! 1! OK
Shift+1:Shift+1 !! !! OK
Shift+1:1 !1 !! WRONG
HTH.
_______________________________________________
Development mailing list
Development@qt-project.org
http://lists.qt-project.org/mailman/listinfo/development
