Re: [orca-list] Punctuation, capital letters, exchange of characters and strings, generally error in the design of Orca

[Willie Walker <William Walker Sun COM>]

Hi Jan:

>     WW> Based upon recent conversations, my current understanding of
>     WW> speech dispatcher is that it merely passes text on to the speech
>     WW> engine with some additional engine-specific markup if the engine
>     WW> supported it.
> Yes.

OK - so, no additional processing of the text is done.  It is just 
dispatched.

>     WW> My argument is that a user wants a feature and they don't care
>     WW> how it is delivered. They just want it to work.
> Mr. Zamazal has already answered. Apart from this we have implemented
> this functionality in Festival through Festival Freebsoft Utils and in
> cooperation with Jonathan Duddington we have achieved that it is
> implemented in Espeak as well.

OK - so some negotiation and work is required from the speech engine 
provider.

>  IBM TTS module in SD emulates much of
> this functionality and if there is need it is possible to find a
> solution for other speech engines as well.

Here's where I get confused.  'Emulates' implies to me that some 
processing is done in SD if the engine doesn't support it.  So, doesn't 
this mean that SD is not just a dispatcher, but it also contains code to 
provide support for features missing in a speech engine?

>     WW> I agree. Let's focus on the user requirements proposed in
>     WW> http://mail.gnome.org/archives/orca-list/2008-March/msg00566.html
>     WW> and discuss how we can deliver these requirements when also
>     WW> dealing with speech engines that do not have direct support for
>     WW> them.
> This discussion has already taken place. You can see the results on
> http://www.freebsoft.org/tts-api . As Mr. Hanke wrote in a previous
> email it is nothing closed and we welcome any suggestions and
> improvements. I don't see any meaning in starting from the very begin
> again.

I'm definitely not suggesting we start from the beginning again.  The 
total number of TTS API's in the world is approaching the total number 
of people in China.

What I'm hearing from you is a proposal that TTSAPI is the solution to 
look at, and that it can be changed to accommodate requirements from 
http://mail.gnome.org/archives/orca-list/2008-March/msg00566.html that 
might be missing.  If that is the case, then a constructive way to move 
forward would be to map the requirements from 
http://mail.gnome.org/archives/orca-list/2008-March/msg00566.html to 
TTSAPI calls.

In addition, I'd also like to hear more about the current state of 
TTSAPI.  How complete is the implementation, where does it work, who is 
using it, etc.

Thanks!

Will