John Levon wrote: > but now they can't see which borders are actually set, regardless of > whether they can change them. A minor thing ...
It is only inside a multicolumn. And there is no border to set/unset in the cases the widget is disabled (apart from column borders which can be toggled from inside a non-multicolumn cell). So, without my color change, you would never ever see the grey color, because it is always unset. I still prefer grey (= no border) as a visual hint to the ambivalent white (= unset border). But that's subjective. > > I have also managed to fix the setborder update problem (a repaint() call > > was needed) > > I think this is wrong, perhaps qsetborder should be calling update() on > itself. Explicit repaints shouldn't be needed... I don't know. I have found this on qt-interest as a hint for similar cases. I have tried a lot of update() and other stuff, neither worked. OTOH, I still do strictly trial and error coding. Jürgen