Re: [orca-list] interesting performance with speech-dispatcher and orca.

[Willie Walker <William Walker Sun COM>]

Hey Tomas:

> Hello, I think I have found the problem.  It seems to be a regression in
> the Orca speech interface, since I haven't seen this in the time of
> writing the backend.

Ha!  Looks like we found the problem at the same time:  the getInfo()
method of speechdispatcherfactory was returning a tuple and not a
list.  

For the record, getInfo() has always been specified to return a list.  I
suspect that the speechdispatcherfactory.py bug wasn't noticed, however,
until we actually started expecting the return value to be a list
elsewhere in the code.

> If it proves to solve the problem for you, I will send a patch to bugzilla.

I put a patch on bugzilla just a couple minutes ago -- we were thinking
along the same lines.  I also just posted your mods as a patch.  Either
one is OK by me.  Can you take a look at them and see which you prefer
since you are the maintainer of speechdispatcher.py:

http://bugzilla.gnome.org/show_bug.cgi?id=349394

> > So folks stay away from SD as speech system until this has been solved. You 
> > can choose SD as synth, that should do no harm.
>
> I thing there is a major misunderstanding going on here.  It has been
> stressed several times that the SD backend is an experimental code.  As
> such it needs to be tested. 

I agree - "use at your own risk" means "use at your own risk".  It
doesn't mean "don't use at all."  :-)

To give you an example of how using this at your own risk has helped:
it has exposed problems, it has brought them to our attention, and we
have at least one potential fix in place.  This community is good stuff.

Will