Summary of Release 5.6

Main New Features

• Provides SSL (https) support, allowing the developer to set any URI in a VoiceXML page to "https://...".
  
  
• New shadow variables have been added to provide detailed grammar information:
  • application.lastresult$.activegrammars[] is an array containing all the grammar objects that were active at the time of recognition.
  • application.lastresult$.triggeredgrammar contains the matched result's grammar object.
  
  
• New shadow variables have been added to provide information about input that terminates a transfer:
  • name$.inputmode is the mode ("DTMF") of the input, if input terminated the transfer; otherwise, undefined.
  • name$.utterance is the dtmf string of the input, if input terminated the transfer; otherwise, undefined.
  Currently, only dtmf termination is supported (no voice termination).
  
  
• DTMF grammar support (both XML and ABNF) is now compliant with the Speech Recognition Grammar Specification (SRGS) Version 1.0, from June 26, 2002.
  
  
• The n best array, application.lastresult$[], (with size <= MAXNBEST), will now be filled with any results, regardless of their confidence level, even when a nomatch event occurs.
  
  
• Added three new attributes: eventexpr, message, and messageexpr, which can be used in <choice>, <link>, and <return>.
  
  
• The platform now accepts application/srgs (ABNF) and application/srgs+xml (XML) as grammar types.
  
  
• The <value> tag will now be supported within SSML, including <emphasis>, <paragraph>, <p>, <prosody>, <say-as>, <sentence>, and <s>.