Re: incorrect signal name leads to python crash

[Tomeu Vizoso <tomeu tomeuvizoso net>]

On Fri, Sep 24, 2010 at 13:02, Barry Warsaw <barry python org> wrote:
> Hi folks,
>
> I'm wondering if anyone has any thoughts on this bug:
>
> https://bugs.launchpad.net/ubuntu/+source/glib2.0/+bug/638513
>
> While the OP fixed his problem by changing the signal name, it still strikes
> me as bad that the incorrect signal name can lead to a Python crash.  I only
> spent a little bit of time debugging this and am not familiar enough with the
> code.  Is this a legitimate bug in pygtk/pygobject?  If so, is there an open
> bug on the issue already (my limited searching on bugzilla did not turn one
> up).

I see this as a consequence of the lack of a way for marking glib API
as not consumable by bindings. gobject-introspection has such a thing
and we'll get eventually.

Even then, mind that unless the Gtk+ developers change their mind
radically, PyGObject won't be able to guarantee that your process
won't crash when passing incorrect arguments. This is because of how
the C implementation has (and lacks) asserts at the start of each
function.

There's some bug in bugzilla about it but cannot find it right now.

Regards,

Tomeu

> Thanks!
> -Barry
>
> _______________________________________________
> python-hackers-list mailing list
> python-hackers-list gnome org
> http://mail.gnome.org/mailman/listinfo/python-hackers-list
>
>