how stable the entire Plasmoid concept is...
I consider Plasmoids as an addition to my usual application. This will allow the user to display part of the information he needs from the application right on the desktop. In Android, this concept of widgets is the basis. In my Linux "KDE Plasma 5" they look very organic too.
It is very convenient to display the temperature of the air conditioner on the desktop, for example. Or have a button for turning on the light, a table lamp. In the age of the Internet of things, this is a promising concept, IMHO.
I think that these plasmoids are for old KDE4.
I found new Plasmoids on the official KDE website:
https://store.kde.org/browse?cat=105&ord=ratingMy standard Linux installation has a lot of Plasmoids out of the box: analog clock, monitoring computer resources (tempo, RAM), network monitoring, sticky notes, bluetooth control, etc.
and they are beautiful

KDE Plasmoids are built on top of Qt's QML, a declarative style markup language to describe both a user interface as well as the event handlers.
At best what Lazarus can do to get involved is writing plugins to interfact with something unreachable from QML level, but the widget must still be written in QML, that's just how Plasmoids are designed.
I began to understand how it works.
It looks like Plasmoids written in Python are essentially QML plugins and a QML interface template.
So, ideally, I need a Plasmoids LCLWidgetSet implementation or a QML-plugin library + an interface template.
But how then is QT implemented in Lazarus? Is there already a translator inside that converts pascal code to QML?
These are probably rhetorical questions and I should investigate this myself...