[cairo] A hidden offset for the Xlib backend

[Jost Boekemeier]

On Sun, 2005-03-13 at 23:32, Owen Taylor wrote:
> > 
> > I like the other approach, making the "device offset"
> > an offset which contributes to the CTM, very much.  If
> > this gets implemented, I could use it, too. (I have to
> > rewrite the code anyway once the new api is there, so
> > this isn't really a problem).
> 
> Can you give more details about your use case?

I can get rid of the cairo_translate when replaying the retained path
onto the surface:

/* compose context */
void
retained_compose_draw (cairo_shape_compose_context_t *_cx, 
		       cairo_shape_drawing_context_t *_cr,
		       cairo_shape_viewport_t *view,
		       cairo_surface_t *surface)

{
  struct retained* cr = (struct retained*) _cr;
  struct retained_compose* cx = (struct retained_compose*) _cx;

  cairo_save(cr->ct); 
  cairo_set_target_surface(cr->ct, surface);
  cairo_translate(cr->ct, -view->bounds.x, -view->bounds.y);
  path_replay(view, cr);
  cairo_restore(cr->ct);
 }


The x/y offset is currently maintained by the library above cairo.  If
we decide to push the offset into cairo, we could call the method
"cairo_set_device_offset_internal()" for example to emphasize that this
method is expected to be called only by the toolkit or the retained mode
immediately above cairo, not by applications.


Jost