Re: [g-a-devel]checking Accessible* validity

[David Bolter <david bolter utoronto ca>]

Thanks Michael:

Michael Meeks wrote:

> Hi David,
>
> On Wed, 2002-12-18 at 20:03, David Bolter wrote:
>
> > Has anyone written (from the at-client perspective) a function to check 
> > the validity of an Accessible*?
>
> 	Nope - there are varying levels of ease with which this can be done.
> Ultimately the Accessible holds a remote reference on the object - so it
> should be sufficient to check the ORB's socket connection between the
> processes to see if the object is still 'alive' - would that suffice for
> you ? [ it would be pretty easy to add some custom code to at-spi that
> would emit a 'defunct' type event on every live at-spi Accessible when a
> client crashed at the remote end - would that help ? ].
That should help actually.  I think it could potentially be a good thing 
for many at-spi clients, but I am not aware of the efficiency costs.

> >  I think most of us clients are queueing and storing such pointers, and 
> > it is possible to cause badness when they are defunct.
>
> 	Badness ? really ? what sort ? you should be able to reliably invoke
> these methods in perpetuity without pathalogical failure - ok so you'll
> get bogus roles [etc.] back, but ...
The example is that gok seg faults (when it tries an Accessible_getName 
([accessible for defunct object]);  I am not sure why it seg faults. 
Note that the process space for gedit is gone at this point.

> 	I'm thinking of caching some of the details [ such as role-type ] so we
> could use that to nicely flag /tag a defunct state perhaps for fast
> polling.
I think I understand you here, but if the cache is finite then there is 
still no gaurantees...

> >  I recently 
> > noticed that gedit does not seem to emit the same defunct events for the 
> > case where it is closed using the file - quit option as opposed to 
> > pressing the close box.
>
> 	Strange.
>
> > You should be able to confirm this using event-listener-test -m  in 
> > at-spi/test
>
> 	And/or at-poke. I think we use children-changed there; are you using
> at-poke [ it crashed for me last time I used it which is sad since I
> spent so long trying to robustify it ;-].
at-poke seems to be working for me...  but does it log events?

> > So given that objects can possibly go defunct without client knowledge, 
> > is there a way a client can check the validity of its stored 
> > Accessible*?  Or is it the case that we must deal with them when we pop 
> > the queue and not store interfaces (e.g. Accessible*) to them longer 
> > than that?
>
> 	My question is - why do you need to know ? :-) also are you using the
> AccessibleEvent_ref / unref things to build your queue ? that's possibly
> the best way to do things in today's world.
I am still using Accessible_ref and Accessible_unref on the 
event->source.  I have been on leave for a few months and am a little 
out of the loop.  Should I be using AccessibleEvent_ref/unref?  Note, I 
am doubtful that will solve this particular issue.

> 	HTH,
>
> 		Michael.

cheers,

David