[orca-list] System Voice patch needs your testing

[Joanmarie Diggs <joanmarie diggs gmail com>]

Hey all.

I've been working with Ale and Juanje on some of the AndalucÃan bugs,
especially those which are in danger of not making it into the next
release due to string and or UI freezes. Reminds me of a quote from one
of my favorite authors, Douglas Adams: "I love deadlines. I love the
whooshing sound they make as they fly by." I wish that GNOME were so
tolerant. But I digress....

Attached to bug 543157 you will find a patch [1] which implements a
system voice, along with refactoring speech generators to provide ACSS
per string (i.e. bug 412656). It could really use some testing and
feedback. 

Given that technically a UI change is being made (addition of an item to
the "Voice settings:" combo box, if we want to get this in for 2.32,
technically it needs to be committed to master no later than Monday.
Therefore, what I'm thinking is that if it works at least "well enough"
and doesn't break anything, we commit it and then refine/tweak as needed
as such tweaks will not introduce UI or string changes.

While we had some initial input on the definition of the purpose of a
system voice, we had to really give it some thought. Having done so, our
initial implementation makes this distinction (or at least attempts to):

* If it's actual text on the screen, it's the default voice
* If it's not actual text on the screen, it's the system voice

Thus the label of a checkbox should be spoken in the default voice. The
role of the checkbox should be in the system voice, as should its
accelerator key, and its state, and the tutorial message. In the case of
a misspelled word announcing 'misspelled' should be in the system voice
because a red squiggly line, while on screen, is not text. I hope this
makes sense -- and I hope that it's at least in the ballpark of what
y'all would expect a system voice to be used for.

By default, the system voice is the same as the default voice. That way,
users who do not like hearing bunches of voices won't notice any change.
But those of you who would find different voices of value, you can
modify the System voice via the Orca Preferences dialog.

Again, if y'all could give this your testing love, with the aim of first
and foremost identifying critical problems which should prevent its
inclusion in Orca, that would be awesome.

Thanks in advance, and our apologies for the short notice.
--joanie

[1] http://bugzilla-attachments.gnome.org/attachment.cgi?id=167897