Hi all,

In a QGraphicsScene, I created a QGraphicsPixmapItem object with no parent and set its transformation origin point to the geometrical centre, then a series of transformations were performed by calling setPos() and setRotation().

The weird problem was that, when I translated the object with
Qt Code: 
  1. setPos(this->scenePos() + delta)
To copy to clipboard, switch view to plain text mode 
, the transformation messed up, but
Qt Code: 
  1. setPos(this->pos() + delta)
To copy to clipboard, switch view to plain text mode 
worked fine.

According to the documentation, pos() is the equivalent to calling mapToParent(0, 0), and scenePos() is the equivalent to mapToScene(0, 0). But the result was not: mapToParent(0, 0), mapToScene(0, 0) and scenePos() returned the same coordinates, while pos() returned a different one.

Totally no idea, what is the coordinate returned by pos() then? Could it be a bug with Qt?

Thanks.