[orca-list] latest AT-SPI2 has an issue (was Re: Have again problems with Gtk Check boxes?)

[Joanmarie Diggs <jdiggs igalia com>]

Hey guys.

After some digging, and some thorough testing from Josà (see below), it
seems that there is an unexpected side effect from a change made to
AT-SPI2. In addition to the Orca key sometimes seeming stuck on when
it's not, and not pressed when it is, I have been able to get my desktop
in situations in which no keys would work at all (e.g. Alt+Tab) until I
physically clicked with the mouse (the actual mouse; not a flat review
click). So.... As you all know, the problem with the bleeding edge is
that it can sometimes cut you. I'm afraid this is one of those times. In
the message below you'll also find a link to the AT-SPI2 patch to
reverse until Mike has a fix.

Many thanks to Josà for looking into this!

Take care.
--joanie

-------- Forwarded Message --------
From: Josà Vilmar EstÃcio de Souza <vilmar informal com br>
Reply-to: vilmar informal com br
To: Joanmarie Diggs <jdiggs igalia com>
Subject: Re: [orca-list] Have again problems with Gtk Check boxes?
Date: Fri, 24 Feb 2012 09:41:12 -0200

Hi Joanie,
I can reproduce the problem all the time and I am using laptop layout.

I reverted the patch as suggested  and now I have good news and I have 
bad news.

The good news are that the problem has gone, and the bad news are that 
the problem has gone!

Seems that the patch is causing some problems.
I tried to reproduce the problem many many many times but the problem 
didn't appear.

To revert the patch I took the folloing steps:

1. Downloaded and saved the content of the patch.
2. Created a new branch, although I think that it isn't necessary.
3.git -R rev.txt # the file where the patch was save.


On 02/23/2012 10:07 PM, Joanmarie Diggs wrote:
> So.... Your debug.out doesn't tell me anything revealing. Plus I am once
> again occasionally seeing this issue myself.<sighs>  Since it comes and
> goes for me, and since I cannot yet reproduce it reliably, I'm wondering
> if you can indeed reproduce it reliably -- and if so, could you try
> reversing this patch to at-spi2-core:
> http://git.gnome.org/browse/at-spi2-core/patch/?id=c5ae7c1b45f9a7156d3a7b5e78c55c7e46ef2ad9
>
> In the meantime, I'll see if I can detect whatever this new condition is
> in Orca and try to prevent it. But any time we hack around stuff in Orca
> that are not Orca bugs we wind up getting bitten down the road when our
> hack breaks something else.
>
> Thanks and take care.
> --joanie