Perhaps gdk_window_create_similar_surface now takes
gdk_window_get_scale_factor into account? /With/ the (second?)
scaling, the surface ink is four times as big as it should be, and
mostly black, except for the appropriately sized image of the blue
ball (with gray background) in its upper-left corner.
There still seems to be a shadow of a (1 pixel?) border around the
surface ink.
(define (get-exposed ink window)
(assert-glib-locked 'get-exposed)
(or (surface-ink-exposed ink)
- (let ((extent (fix-ink-extent ink))
- (scale (C-call "gdk_window_get_scale_factor" window)))
- (let ((width (fix:* scale (fix-rect-width extent)))
- (height (fix:* scale (fix-rect-height extent))))
+ (let ((extent (fix-ink-extent ink)))
+ (let ((width (fix-rect-width extent))
+ (height (fix-rect-height extent)))
(let* ((surface (gdk-window-create-similar-surface window
width height))
(cr (cairo-create surface)))