The voiceglue project is an attempt to create a
free implementation of VoiceXMML 2.0 with VoiceXML 2.1 additions.
Our goal is to first implement all of the required
features of the 2.0 specification http://www.w3.org/TR/voicexml20/
and 2.1 additions http://www.w3.org/TR/voicexml21/
though several optional features
have already made it.
As of voiceglue 0.12, we have not yet achieved
a full 2.0 implementation.
But we're getting close.
This is not an exhaustive list, but hits most
of the major points.
Full
HTTP document fetching, parsing, and syntax support
Full javascript (ECMAscript) processing
Value passing to and from Asterisk
Cookie support
Pre-recorded
HTTP audio fetching and output
TTS output
Audio recording and retrieval
DTMF input with simple SRGS
XML grammars
Primitive blind transfer support
The data tag from VXML 2.1
These are the major required VXML 2.0 features
not present in voiceglue 0.12:
Support for Ubuntu Karmic and Lucid
Support for recent RHEL and CentOS
Increase SpiderMonkey memory allocations
Added session.connection.local.uri and …remote.uri
Fixed race condition bugs
Fixed memory leaks
Removed delay-inducing OpenVXI caching
Added ssml_passthrough option in voiceglue.conf
Added language parameter to TTS interface
Added passing variables to and from Asterisk
