-------------------------------------------------------------------- 
 copyright 2001 by voxeo corporation. (see LGPL.txt) 
-------------------------------------------------------------------- 

 v1.0 coded in ASP, ColdFusion and PHP for VoiceXML
 by Kathryn Sharron 

This concise application tells knock-knock jokes.

The zip file contains the following files: 

  knock_a.asp   an ASP document utilizing audio files
  knock_db.asp  an ASP document utilizing TTS from a database 
  knock_f.asp   an ASP document utilizing TTS from a text file 
  knock.cfm     a ColdFusion document utilizing TTS from a database 
  knock_a.cfm   a ColdFusion document utilizing audio files
  knock.php     a PHP document utilizing TTS from a text file
  knock_a.php   a PHP document utilizing audio files

  knock.grammar     grammar file
  knock_a.grammar   a grammar file for the voicexml documents 
                    that use audio files 

  Knock.mdb   an Access database of jokes
  knock.txt   a text file of jokes

  LGPL.txt     opensource license agreement 
  Readme.txt   Introduction to this application 

and in the subdirectory 'audio':

  KNOCK.wav      "Knock knock!"
  AGAIN.wav      "Would you like to hear another?"
  THANKYOU.wav   "Thank you."

  SETUP1.wav ... SETUP20.wav       i.e., "Cows" (cows who?)
  DELIVER1.wav ... DELIVER20.wav   i.e., "No, cows moo"

Endpointing (ep) Parameters:
  
  nuance.core.ep.EndSeconds
  Type: float  
  Default: 0.75  
  Minimum: 0.20  
  Maximum: 10.00
  
  Specifies the minimum amount of silence required to indicate 
  that the end of speech has occurred. A longer value is less 
  likely to cut the talker off, but also makes the interaction 
  slower. This parameter is often set from within applications, 
  based on the type of response expected from the user. For 
  example, 0.5 seconds might be appropriate for a simple yes/no 
  question, whereas 1.5 seconds might be more appropriate when 
  a talker is likely to pause mid-sentence.

  nuance.core.ep.PrepareForBargeIn
  Type: bool  
  Default: false
  
  Set to TRUE if your system includes barge-in to allow the 
  endpointer to provide better performance. This enables the 
  endpointer to better detect start of speech that occurs 
  during prompt playback. 

  nuance.core.ep.ThresholdSnr  
  Type: int  
  Default: 10  
  Minimum: 5  
  Maximum: 30
  
  Controls the threshold of noise the endpointer needs to detect 
  start-of-speech and end-of-speech. To reduce the likelihood of 
  false detection in noisy environments, you can increase the 
  value of this parameter. However, this requires the user to 
  speak more loudly to be heard. Typically you never need to 
  modify this parameter. Increasing the value of this parameter 
  makes the application less sensitive to barge-in. 
  
*TTS = text to speech

--------------------------------------------------------------------  