Hi Denis
I am very interesting in GTK+ and DirectFB. When I ported the Gtk+
with DirectFB, the FullScreen seemed not support. Will it be a
feature.
My get paid work is develping DirectFB and writing some gfx drivers.
After cross-compiling Gtk+ with DFB in arm11, I noticed that the gtk+
programs doesn't use the drawing functions but blitting only. The
performance was not good to apps with scrollbars. Is it normal or not.
My envirnoment is:
GTK+ with 2.10
DFB with 1.0.0
2008/1/30, Denis Oliver Kropp <[EMAIL PROTECTED]>:
> Hi,
>
> this is the first of more patches coming up these days.
>
> It fixes various warnings, some errors, has some cleanups and a change that's
> worth to mention:
>
>
>
> @@ -2818,10 +2818,10 @@
> }
>
> static void
> -gdk_window_impl_directfb_invalidate_maybe_recurse (GdkPaintable *paintable,
> - GdkRegion *region,
> - gboolean (*child_func)
> (GdkWindow *, gpointer),
> - gpointer user_data)
> +gdk_window_impl_directfb_invalidate_maybe_recurse (GdkPaintable
> *paintable,
> + const GdkRegion *region,
> + gboolean
> (*child_func) (GdkWindow *, gpointer),
> + gpointer
> user_data)
> {
> GdkWindow *window;
> GdkWindowObject *private;
> @@ -2866,14 +2866,11 @@
>
> if (child_func && (*child_func) ((GdkWindow *)child, user_data))
> {
> - gdk_region_offset (region, - child->x, - child->y);
> - gdk_region_offset (&child_region, - child->x, - child->y);
> gdk_region_intersect (&child_region, region);
> + gdk_region_offset (&child_region, - child->x, - child->y);
>
> gdk_window_invalidate_maybe_recurse ((GdkWindow *)child,
> &child_region,
> child_func, user_data);
> -
> - gdk_region_offset (region, child->x, child->y);
> }
>
> temp_region_deinit( &child_region );
>
>
>
> In case I'm not missing something here, this change does not only fix the
> conflict with
> the new const, but also reduces 4 calls to one (gdk_region_offset).
>
>
> Mike was asking for someone to take over the project recently, wasn't he?
>
> After seven years of more and more abandoning it, I'd be happy to have it
> back and make
> it something great(er) :)
>
> Too bad noone could tell me about the simpler Quartz implementation so far.
> It seems a
> lot of the recursion with invalidation, clipping, lots of paint events etc.
> could be
> avoided. There are also other differences like, e.g. crossing event
> generation does not
> happen on show/hide of windows.
>
> --
> Best regards,
> Denis Oliver Kropp
>
> .------------------------------------------.
> | DirectFB - Hardware accelerated graphics |
> | http://www.directfb.org/ |
> "------------------------------------------"
>
> _______________________________________________
> directfb-dev mailing list
> [email protected]
> http://mail.directfb.org/cgi-bin/mailman/listinfo/directfb-dev
>
>
>
_______________________________________________
directfb-dev mailing list
[email protected]
http://mail.directfb.org/cgi-bin/mailman/listinfo/directfb-dev
