Re: [evolution-patches] gal, #59047, Gnopernicus can't read evolution's message list properly

[Eric Zhao <eric zhao sun com>]

Hi Michael,

The eti_a11y_cursor_changed_cb function will only be invoked in the
following situation:
(1) some cell in the table is clicked and eti_event is invoked.
(2) eti_event function will use find_cell (eti, realx, realy, &col,
&row, &x1, &y1) to find the clicked cell's row and col. 
(3) Later the row and col will be converted to model format and
e_selection_model_maybe_do_something will be invoked.
(3) In e_selection_model_maybe_do_something , "cursor_changed" signal
will be emitted and eti_a11y_cursor_changed_cb will be called. 

  Since only the visible cell can be clicked, no assertion will hitted
in atk_table_ref_at. Hope I have expressed clearly.

Regards,
Eric.

   
On Wed, 2004-05-26 at 15:14 +0800, Not Zed wrote:
> 
> Should this handle as a normal case, when the model row isn't shown,
> i.e. it has no view row/column?
> 
> Otherwise it will only 'work' by hitting an assertion in
> atk_table_ref_at.
> 
> 
> On Tue, 2004-05-25 at 22:40 +0800, Eric Zhao wrote: 
> > Hi,
> >    #59047 is an a11y related bug. The bug's detailed information and
> > reason is at http://bugzilla.ximian.com/show_bug.cgi?id=59047 .
> > 
> >    I provide a patch to fix this bug, please help review, thanks.
> > 
> > Regards,
> > Eric
> 
> Michael Zucchi
> <notzed ximian com>
> 
> Ximian Evolution and
> Free Software Developer
> 
> 
> Novell, Inc.