Speech-dispatcher c api questions

[Jeremy Whiting]

On Tuesday 16 June 2009 06:58:33 Hynek Hanke wrote:
> Jeremy Whiting napsal(a):
> > 2) I'm using one connection for the whole lifecycle of kttsd (unless it
> > goes down, then I reconnect.  Is this the preferred method? or should I
> > be making a new connection for each speech job?
>
> I think you must carefully think this through. How independent are
> the KTTSD clients? Is it desirable to be able to set them speech parameters
> (such as synthesizer, voice and volume) independently? If KTTSD is
> kind of a relay between the app and Speech Dispatcher, then the
> Speech Dispatcher way is to have separate connections for separate clients.
> If on the other side KTTSD is more like the mouth for KDE which looks here
> and there and speaks to the user about what is happening, than a unified
> approach
> with one connection could be prefered.
>
> > 3) When I set kttsd up to send all irc notifications through
> > speech-dispatcher I notice that jobs are getting truncated.  I'm
> > currently saying all jobs with a text priority and it seems when a new
> > job comes in, the previous job is cancelled and the new one started.
>
> This is the correct behavior. See
> http://cvs.freebsoft.org/doc/speechd/ssip_10.html#SEC11 .
> Looks like you nead priority MESSAGE instead.

defaulting to Message works much better, yes, thank you =)

> > 4) Is history completely unimplemented? or just not supported in the c
> > api? I've noticed a few calls return error codes instead of valid
> > information, such as the history get message list command.
>
> History is not really working neither on server side. Something is
> implemented,
> but not finished and likely quite obsolete by now. SSIP commands should
> return an error. There was never any real demand for history, so we didn't
> work on it much. Perhaps for accessibility purposes it's better to
> reproduce the situation (e.g. by cursor navigation) than to try to find an
> old message in a pool of many other old messages.

That makes sense, I was just thinking of making kttsd's job manager basically 
just wrap speech-dispatcher's history, but that might be useless, I'm not 
sure.  If kttsd users ask for job control I'll give some time/thought to 
implementing history in speech-dispatcher.

> > 5) Hynek, I've noticed a couple inconsistencies between the c api and the
> > c api's documentation, should I send you notes directly about those? or
> > will they be fixed automatically next time you run the texi2html or
> > whatever generates the documentation?
>
> The version of documentation from www.freebsoft.org used to be
> updated automatically from CVS till yesterday, when we moved to GIT.
> So no, you will likely have to send patches to fix those inconsistencies.
>
> Have a nice day,
> Hynek

In the C API documentation the last method (spd_list_synthesis_voices) has the 
wrong return type here: http://cvs.freebsoft.org/doc/speechd/speech-
dispatcher_14.html#SEC56 it actually returns an SPDVoice** not a char** =)

Jeremy