Re: [evolution-patches] gal e-table patches

[Li Yuan <Li Yuan Sun COM>]

JP Rosevear wrote:

> On Tue, 2005-01-25 at 23:37 -0500, JP Rosevear wrote:
>  
>
> > On Thu, 2005-01-20 at 12:19 +0800, Li Yuan wrote:
> >    
> >
> > > Hi, JP
> > >
> > > Here are our e-table a11y patches.
> > > Please help me to review it.
> > >      
> > Ok, so in general it looks ok, couple points:
> >
> > e-cell-toggle.c:  Were you actually hitting an error that caused you to
> > add this range check?
> >    
It is a old internal bug which will cause hydrogen crash. If you do not 
want it, I can remove
it form the patch.

> > e-table-click-to-add.[hc]: It looks like you started to add a style_set
> > signal but then realized you only needed to set a vmethod but did not
> > remove the other bits
> >    
Because click-to-add is a GnomeCanvasItem in e-table->canvas_vbox. And 
canvas_vbox is in e-table->table_canvas,
which is a e-canvas.
when style changed, e_canvas_style_set calls ec_style_set_recursive. it 
will look for the "style_set"
single in GnomeCanvasItem.

> > e-table-subset.c: This one scares me a bit, if this is indeed sent out
> > wrong why hasn't this bitten us before?
> >    
It scared me a bit too. The function is used to connect 
"model_cell_changed" signal.
And in e-table-model.h, the definition of the function is

void        (*model_cell_changed)  (ETableModel *etm, int col, int row);

> One more thing, the Changelog should contain the bug numbers of the
> fixes, in context, ie:
>
> Fixes #11111
> a11y/e-a11y-table.c: <description>
>
> Fixes #22222
> a11y/e-a11y-click-to-add.c: <description>
>
> This makes it easier to tell what each bit is supposed to fix and makes
> it easier to collect up release note information.
>  
ok, I will add the number to the Changelog

> -JP
>