W3C home > Mailing lists > Public > www-voice@w3.org > July to September 2008

RE: FW: Vxml grammar with termtimeout=0s. How should the intrepeter react?

From: Michel Brabants <Michel.Brabants@telindus.be>
Date: Wed, 24 Sep 2008 16:10:20 +0200
Message-ID: <4C95A2806C1B5C42BBC5801B08691501027C476E@TBE1MXCV1.telindus.intra>
To: "David Nicol" <davidnicol@gmail.com>
Cc: <www-voice@w3.org>

Hey David,

Thanks for the help. I'm working with support at the moment.

Kind regards,


-----Original Message-----
From: David Nicol [mailto:davidnicol@gmail.com] 
Sent: 23 September 2008 20:12
To: Michel Brabants
Cc: www-voice@w3.org
Subject: Re: FW: Vxml grammar with termtimeout=0s. How should the
intrepeter react?

On Mon, Sep 22, 2008 at 8:48 AM, Michel Brabants
<Michel.Brabants@telindus.be> wrote:
> Hello,
> (2nd time I'm sending because the first one seemingly didn't arrive.)
> I've written a small vxml-grammar and the product on which we use it 
> is not reacting like I think it should. Could someone confirm or 
> correct what the expected behaviour should be according to vxml 2.1?

the spec says: "If the termtimeout is 0s (the default), then the
recognized value is returned immediately after the last DTMF allowed by
the grammar, without waiting for the optional termchar. Note: the
termtimeout applies only when no additional input is allowed by the
grammar; otherwise, the interdigittimeout applies."

It appears to me that your expected behavior is correct, and the product
you are using is out-of-spec.  I would escalate this to their support.
Grammar parsers are tricky things, and they probably aren't testing for
the corner case you present.  For instance, their parser might think
that up to eleven keypresses are allowed and send a lower layer a
request for up to eleven keypresses with a 3 sec timeout, then work with
what comes back, rather than requesting digits as they appear.

If you revise to use a termchar instead of expecting the hardware (which
might be able to accept and respond to requests like "give me N
keypresses, with a T timeout" but nothing any more nuanced than that) to
understand a three-pronged state machine, do you get the results you
Received on Wednesday, 24 September 2008 14:11:30 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:07:40 UTC