----- Original Message -----
> > + gint xr, yr;
> >
> > - gint xr = gdk_screen_get_width(screen) / 2;
> > - gint yr = gdk_screen_get_height(screen) / 2;
> > -
> > +#ifdef WIN32
> > + RECT clip;
> > + g_return_if_fail(GetClipCursor(&clip));
> > + xr = clip.left + (clip.right - clip.left) / 2;
> > + yr = clip.top + (clip.bottom - clip.top) / 2;
> > +#endif
> > if (xr != (gint)motion->x_root || yr != (gint)motion->y_root)
> > {
>
> xr and yr are no longer initialized at this point on !WIN32
oops, indeed! I have a bit too much of a windows centric eye these days..
> Patch looks good otherwise.
I am reworking the patch, since it doesn't work when the viewer is in !primary monitor. Something like that:
- if (!SystemParametersInfo(SPI_GETWORKAREA, 0, &workarea, 0))
+ monitor = MonitorFromRect(&window, MONITOR_DEFAULTTONEAREST);
+ g_return_val_if_fail(monitor != NULL, false);
+
+ mi.cbSize = sizeof(mi);
+ if (!GetMonitorInfo(monitor, &mi))
goto error;
Unfortunately, I don't have dual-head physical windows machine atm to really test this patch, but it still seems to fail in some corner-cases..
thanks
_______________________________________________
Spice-devel mailing list
Spice-devel@xxxxxxxxxxxxxxxxxxxxx
http://lists.freedesktop.org/mailman/listinfo/spice-devel