(In reply to alexander :surkov from comment #35)
> (In reply to Trevor Saunders (:tbsaunde) from comment #34)
> > Comment on attachment 726580
> > Implement a replacement of atk_object_set_name() which mimics the behavior
> > without calling atk_object_get_name()
> >
> > >+static void
> > >+AtkObjectSetName(AtkObject *aAtkObj, const gchar *name)
> >
> > so, both of the times this is called we have to check if we actually want to
> > fire an event first.
>
> we don't want an event in the former case
not sure what you mean. both places we call atk_object_set_name() we guard it with something that is more or less oldName != newName
now, the first of those shouldn't be needed in an ideal world because we'd always fire name change event, but we don't do that because its between hard and impossible to do that because name algorithm is complicated and depends on all sorts of stuff.
(In reply to alexander :surkov from comment #35) set_name( ) which mimics the behavior get_name( ) ame(AtkObject *aAtkObj, const gchar *name)
> (In reply to Trevor Saunders (:tbsaunde) from comment #34)
> > Comment on attachment 726580
> > Implement a replacement of atk_object_
> > without calling atk_object_
> >
> > >+static void
> > >+AtkObjectSetN
> >
> > so, both of the times this is called we have to check if we actually want to
> > fire an event first.
>
> we don't want an event in the former case
not sure what you mean. both places we call atk_object_ set_name( ) we guard it with something that is more or less oldName != newName
now, the first of those shouldn't be needed in an ideal world because we'd always fire name change event, but we don't do that because its between hard and impossible to do that because name algorithm is complicated and depends on all sorts of stuff.