Re: GnomeClient replacement?

[Bill Haneman <Bill Haneman Sun COM>]

Looks to me as though this has caused a major regression, if I
understand correctly.

Back in the ancient days when the AT support was initially being added,
we were discouraged from relying on environment variables to control the
loading of accessibility modules.  The gconf and gnome_program_init
stuff was written to address this "problem".  Why has the recommended
solution suddenly become an 'ugly hack' ?

The GTK_MODULES solution has some nasty issues - some theoretical, some
real - for instance it tends to break gtk-1.2 stuff badly.  It also
means that we have to load gnome-specific code into gtk-only programs if
we don't use gnome_program_init, since there's no way to tell whether
libgail-gnome needs to be loaded or not; we just have to add it to the
GTK_MODULES list.  (If nothing else, we need to make this change in
gnome-session, in order to reverse the regression).

A better solution would be to use an XSETTING for a11y instead of just a
gconf key, so that apps could detect it w/o a gconf dependency.  This
has been discussed in the past and seems to me it would have been better
than ripping the gnome_program stuff out first.  As part of this, gtk+
would need to load the appropriate modules at runtime _itself_; i.e.
every app and/or toolkit would be responsible for reading the "please
turn on AT support" key and loading the appropriate modules.  In the
case of gtk+ apps, the modules required are libgail and libatk-bridge;
for gnome apps that use embedding and bonobo, libgail-gnome is also
required.  Some apps load their own specialized code as well, for
instance firefox and openoffice.

Bill

On Tue, 2006-07-25 at 16:29, Elijah Newren wrote:
> On 7/25/06, Bill Haneman <Bill Haneman sun com> wrote:
> > Because of build issues I have not been able to test a11y with
> > gnome-session HEAD.
> >
> > Are you telling me that a11y is regressed severely in HEAD, as it
> > sounds?
> 
> No, I'd have to know whether it was in order to say so.  I just
> finally caught up on part of my email and since Havoc was bringing up
> a bug from a few months ago (and made it sound less safe than I
> thought it was at that time), so I thought I'd comment.
> 
> > GTK_MODULES alone will NOT work for the gnome desktop as a whole, since
> > libgail-gnome should be selectively included.
> >
> > Elijah, why didn't you take this to the accessibility-devel-list?
> 
> accessibility was cc'ed on the bug (okay, a different alias but
> probably close, right?), and my reading of the bug said that it should
> be closable as soon as the blocking gtk+ bug was closed.  It was
> closed, and had been for a long time.  So, I added a patch and waited
> for about two weeks for comment, then just went and committed.  I
> really didn't think it'd cause any adverse affects, at least not until
> I read Havoc's email yesterday.  After that I wasn't so sure anymore.
> 
> *shrug*
> 
> Elijah