Re: [orca-list] pidgin 2.6.4 and orca

[Joanmarie Diggs <joanmarie diggs gmail com>]

Hi Will, Zack.

On Tue, 2009-12-01 at 09:44 -0500, Willie Walker wrote:

> I haven't had a chance to test pidgin 2.6.4.  Unlike the stock market, 
> past performance from the pidgin team is usually a good indicator of 
> future performance, which means they are likely to break accessibility 
> when they update the user interface. ;-)

Heheh. In this case, having just built pidgin 2.6.4 in OpenSolaris, it
looks like they may have changed something that we were counting on
w.r.t. the hierarchy and/or events which get emitted. Using Accerciser,
I'm seeing appropriate events and accessible objects. In light of
this...

Zack, I'd ask that you'd hold off on the following:

> Can you please log a bug with the Pidgin project and CC me?  I can 
> certainly do it, but I think the message is much more powerful if it 
> comes from an end user.
>
> http://developer.pidgin.im/wiki/TipsForBugReports

At least until further triage indicates that they really did break
something.

What is helpful when stuff like the following is observed:

> Zachary Kline wrote:
> > Hi All,
> > I made the rather unpleasant discovery today that ArchLinux upgrading
> > pidgin to 2.6.4 broke the orca script for it completely.  I have orca
> > 2.28.1 at the moment from the arch repository.

is to file a bug against Orca in GNOME's bugzilla. That ensures it
remains on our radar. I've gone ahead and filed one for you:
https://bugzilla.gnome.org/show_bug.cgi?id=603520

The other thing which would be helpful -- phenomenal, even -- is some
community help with bug triage. For instance, if the bug is "Orca no
longer speaks incoming chat messages in Pidgin" (which is what I'm
seeing with 2.6.4), Accerciser's event monitor and/or Orca's debug
output could be used to see if those messages appear in the form of
text-changed events. If they are (and it seems that they are), then it's
likely an Orca bug. If they are not, then it's likely a Pidgin bug.

Where this comes in handy, at least for me, is when a little spare time
presents itself. Today, I took a lunch break from my day job and decided
to look at this issue. The time spent doing the triage, filing the bug
against Orca, and composing this message is about the same amount of
time I suspect it would've taken me to fix a problem of this nature.
Unfortunately, my lunch break is now over, so I didn't actually get to
working on the fix.

The other thing that would come in handy is more specific details about
what has broken. When I saw "broke the orca script for it completely" I
was expecting some serious and disturbing accessibility flaw. In my
testing thus far, that has not proven to be the case. If you're seeing
additional problems with the latest version of Pidgin, please add them
to the above Orca bug.

> > I wonder if this sort of breakage is reason enough to run orca from the
> > source repository?

That wouldn't help you with this particular bug -- at least not yet --
because we didn't know about this particular bug. Once we've addressed
this bug, using git master would be the way to get the fix immediately.
Otherwise, it should wind up in the next release of 2.28.

Thanks and take care.
--joanie