RE: The <nomatch> element inside a <record> element.

I understand that provision of speech recognition during recording is
allowed by VXML 2.0 therefore presumably this is when a nomatch might occur.

btw: This comes from a spec reader not coder.

-----Original Message-----
From: Scott McGlashan [mailto:scott.mcglashan@pipebeach.com]
Sent: Friday, 20 September 2002 8:15 a.m.
To: Hoban, Thomas X.; www-voice@w3.org
Subject: RE: The <nomatch> element inside a <record> element.



Hi Tom,

I agree -- there is no situation I can see in which a nomatch event
would be thrown during recording. However, the app developer could add a
<filled> inside <record> and then <throw> a nomatch event from there. 

In VoiceXML 2.0, when the recording exceeds maximum duration, recording
is terminated and the maxtime shadow variable is set to boolean true. On
a historical note, there is nothing in VoiceXML 1.0 which specifies that
a nomatch should be thrown if maxtime is reached (the simplest reading
of that spec for me is that the recording would terminate normally when
the maxtime is reached). 

Scott


-----Original Message-----
From: Hoban, Thomas X. [mailto:thoban@shoptalk.com]
Sent: Thursday, September 19, 2002 20:30
To: www-voice@w3.org
Subject: The <nomatch> element inside a <record> element.



In the newest VXML spec, it appears to me that <nomatch> is a valid
element inside of a <record> element.  However, I don't see anything in
the specification that describes what would cause a nomatch event to be
thrown while recording.  In one VXML 1.0 product that I'm working with,
I see nomatch thrown when the recording exceeds the maximum duration.

Can anyone give me insight into what should cause a nomatch while
recording?

Thanks
Tom Hoban

Received on Monday, 14 October 2002 19:04:22 UTC