gtkmm2.4 and MSVC (again)

[m h <msh126 yahoo com>]

i've been able to trace the problem to a potential
source.
the problem i've been having with the DLLs that were
buildable a few weeks ago was the source code
generated by glade--.
the basic model: class A inherits from Gtk::Window,
class B inherits from class A.  any member variables
in class B caused a non rtti object exception.
after tracing thru gtkmm, glibmm, gtk, glib, back to
gtkmm, it seems the problem occurs on a dynamic_cast
(not a surprise) on the Widget_class callbacks
(show_callback in my case).
the Glib::ObjectBase::_get_current_wrapper call fails.
i'm assuming the point is to take a GtkWidget and get
the actual object (in my case class B) that contains
the widget.
looking briefly thru this code, it hints at something
i've been suspicious about the whole time.  the
static/global variables Glib::quark_,
Glib::quark_cpp_wrapper_deleted_, and
Glib::glibmm_null_pointer.
i've always gotten warnings with these.  some multiple
definitions, some dll-inconsistent linkages, etc.
any idea on how to fix these?
could these be causing the problem with not being able
to find the wrapper?
thanks
--Matt


	
		
__________________________________
Do you Yahoo!?
Yahoo! Mail - You care about security. So do we.
http://promotions.yahoo.com/new_mail