> So this a bug in the vga/pcdisplay driver? In principal, yes. But I simply don't have time to rewrite the soft cursor code, and this works around the problem without impacting anything else.