- From: James Larson <jim@larson-tech.com>
- Date: Tue, 15 Apr 2008 22:10:22 -0700
- To: www-voice@w3.org
The W3C Voice Browser Working Group held its quarterly face-to-face meeting March 5-7 in Orlando. We thank Voxeo for hosting our meeting, giving us a tour of their data facility, providing us a nice working environment, and hosting activities outside of the meeting including a tour of the Kennedy Space Center. During much of the face-to-face meeting, we reviewed preliminary draft sections of VoiceXML 3.0. VoiceXML 3.0 will have at least two profiles: Media Server Profile and VoiceXML 2.1 Profile. Applications written as 2.1 documents can be executed under VoiceXML 3.0 processor using the 2.1 profile. As an example, the Implementation Report tests for VoiceXML 2.1 (which include the Implementation Report tests for VoiceXML 2.0) will be supported on VoiceXML 3.0 platform. There are some exceptions: the VoiceXML 2.1 specifications will be tightened to clarify ambiguities and improve interoperability. Specifically, the built-in grammars will no longer be supported, and several parameters with “platform dependent” default values will be assigned specific default values. We discussed the new <media> element, which is an enhanced and generalized version of the VoiceXML <audio> element. It will permit media formats which contain audio and video tracks. The <par> element will specify playback of media in parallel, while <seq> will specify playback in sequence. We discussed the three new VoiceXML 3.0 functions: We discussed following three new VoiceXML 3.0 functions, and also discussed how they relate to speech recognition and speech prompting: * speaker enrollment * speaker identification * speaker verification and how they relate to speech recognition and speech prompting. We also reviewed and clarified our requirements for recording speech. In non-VoiceXML 3.0 activities, we reviewed and discussed some errata for VoiceXML 2.0/2.1 and voted to publish another working draft of SSML 1.1. Our next face-to-face meeting is scheduled for June 9-11, 2008 in Saarbrueken, Germany. Jim Larson Voice Browser Working Group Co-chair
Received on Wednesday, 16 April 2008 05:11:00 UTC