Re: grab_default doesn't seem to grab.

[csincock-gmail <csincock gmail com>]

perhaps this 'fix' from 1999 is causing the problem
http://mail.gnome.org/archives/gtk-app-devel-list/1999-January/msg00134.html

anyway I ran your script and it doesn't work for me either.
also it seems that gtk is lying about the whole situation.
I added print statements:
print STDERR $close_button->has_default;
print STDERR $close_button->receives_default;
and it reports them both as true.

what's more, calling $window->activate_default after the window is realised
successfully sends a 'activate' signal to the close button, but neither 
buttons clicked handler is called (I guess cos 'activation' is 
different than clicking if the activate is triggered programatically).

Also, if both buttons have $button->can_focus(FALSE)
then the enter key seems to succesfully invoke the close buttons activate
signal handler and it's clicked signal handler.

perhaps having 'focus' support and a 'default' support are 
mutually exclusive in the situation where multiple widgets
accept an enter key when focused?

I think the only way you'll get it to work is by also playing 
around with the focus.

having said that, perhaps just focusing the close button is the way
to go. alternatively you could also add a key event handler to the toplevel 
widget to intercept the enter key and call window->activate if you also
connected your close handler to 'activate'.

I agree with you that the behaviour is not what the docs would seem
to indicate, but because of the interactions of focus handling etc,
I wouldn't be surprised if anybody in the world could say how it is
really "meant" to work. If you collated all the hundreds of single-liner 
comments in the docs for the various focus and 'default' related features, 
you'd probably find plenty of contradictions.

cheers
> Date: Fri, 29 May 2009 14:27:55 +0200
> From: Cowley Harris <warewolfe gmail com>
> Subject: grab_default doesn't seem to grab.

> Trying to understand why grab_default doesn't seem to work like the
> way I interpreted the documentation.
> Can somebody run this script and see if the issue is explicable ?