VoiceXML architecture question

Hello!

I'm new in VoiceXML and I try to understand the architecture of an environnement for vocal services running with VoiceXML.

Maybe, someone could answer to my questions?

What I understand is that it is the role of the implementation platform to play vocal messages. This vocal messages are loaded from a web server (maybe http server?) by the interpreter.
Is it correct?

So does it means that the implementation platform should have (/be linked to) a http client that contains a cache ?

Indeed, downloading audio files could take a lot of ressouces and so the user may hear silence. 
Does VoiceXML forcast mecanisms for provisionning? 

Thanks,

Laurent.
 

----- La messagerie itinérante sans abonnement NetCourrier -----
Web : www.netcourrier.com - Minitel : 3615 NETCOURRIER
Téléphone : 08 36 69 00 21

Received on Thursday, 8 February 2001 12:05:38 UTC