Re: Unsupportable ATK interfaces??

[Bill Haneman <Bill Haneman sun com>]

>Hi Bill,
>
>Currently I am looking at the following approach after feedback from 
Peter 
>and Padraig:
>AtkInterfaces that aren't going to be implemented/relevant will just be 
>ignored so iface->function_name  function pointer will be a NULL *. 
>If the data returned is to be trivial or meaningless then why bother 
>even writing code for it?

OK - this works because the ATK wrapper functions for the function 
pointers currently "do the right thing" if they encounter a null 
function pointer.

>One point of concern about the GtkClist widget is that it was a common
>sight in Gtk+1.2/Gnome1.4 applications: This will be an issue when
>existing 1.4 gnome applications are to be ported to use the Gtk+2.0 
API.
>Since it's not absolutely necessary many component owners might not 
bother
>reimplementing GtkClist's as GtkTree's

Yep.  This is a concern, I hope we can implement the most essential 
parts of the API for GtkCList.

-Bill

>Thanks,
>
>Niall
>

------
Bill Haneman x19279
Gnome Accessibility / Batik SVG Toolkit
Sun Microsystems Ireland