|
| Disconnect call after 120 seconds of inactivity |
 |
Wed, 18 Apr 2007 09:48:33 -040 |
I am having a problem when the user makes a mistake and taps the ringer hook
and does not fully hang up and terminate the call. They then get a stutter
dial tone (beeping dial tone) and dial the applications number again which
transfers the call to another instance of the application. This causes at
least one channel to be tied up until midnight when the server recycles.
This is reducing throughput since a channel is lost every time this happens.
I allow the users to press * to start over but I don't think everyone is
doing this.
I had posted about this a while ago and got a response to set a timer in
javascript to disconnect the call but I am having trouble getting this to
work because setTimeout() does not appear to be supported in whatever
version of javascript the Speech Application SDK 1.1 uses.
Has anyone run into this problem or does anyone have any ideas on how to
disconnect the call in this situation. Simply disconnecting after 2 minutes
would be fine as I am sometimes seeing calls at 15 hours long! I need to
roll another application out and would like to have this problem resolved
before I do.
Thanks in advance!
Andy
|
| Post Reply
|
| Re: Disconnect call after 120 seconds of inactivity |
 |
Mon, 25 Jun 2007 14:36:36 -040 |
I believe I have finally found a solution to this problem and as I thought,
it was very simple.
The solution is in the SilenceAndLowConf sample in C:\Program
Files\Microsoft Speech Application SDK 1.1\Applications\Samples
To make sure the call is ended I added a CatchThreeMumblesQA QA control
right after my initial Welcome QA that says Welcome once. This QA has the
following ClientActivation script and OnClientComplete script which will
play any QA prompt a max of three times and then termminate the call. I
found out that window.close() ends the call which I also didn't know. I
also unchecked BargeIn and checked PlayOnce.
// Use the client activation function of the first QA on the page to count
the number
// of consecutive silences and/or mumbles. By putting the logic into a
separate QA,
// we are able to do the counting in one place, rather than in each QA
separately.
function CatchThreeMumblesQA_ClientActivationFunction(theQA)
{
if (RunSpeech.ActiveQA == null) return(false);
var h = RunSpeech.ActiveQA.History;
return((h.length > 2) &&
((h[h.length-1] == "Silence") || (h[h.length-1] ==
"NoReco")) &&
((h[h.length-2] == "Silence") || (h[h.length-2] ==
"NoReco")) &&
((h[h.length-3] == "Silence") || (h[h.length-3] ==
"NoReco")))
}
function CatchThreeMumblesQA_OnClientComplete()
{
// We call window.close, which will disconnect the call (and thus no more
prompts will be played)
window.close();
}
Let me know if you have any questions.
Thanks!
Andy
"Andy Klare" <klar02@hotmail.com> wrote in message
news:uTdvKBcgHHA.4416@TK2MSFTNGP03.phx.gbl...
>I am having a problem when the user makes a mistake and taps the ringer
>hook and does not fully hang up and terminate the call. They then get a
>stutter dial tone (beeping dial tone) and dial the applications number
>again which transfers the call to another instance of the application.
>This causes at least one channel to be tied up until midnight when the
>server recycles. This is reducing throughput since a channel is lost every
>time this happens. I allow the users to press * to start over but I don't
>think everyone is doing this.
>
> I had posted about this a while ago and got a response to set a timer in
> javascript to disconnect the call but I am having trouble getting this to
> work because setTimeout() does not appear to be supported in whatever
> version of javascript the Speech Application SDK 1.1 uses.
>
> Has anyone run into this problem or does anyone have any ideas on how to
> disconnect the call in this situation. Simply disconnecting after 2
> minutes would be fine as I am sometimes seeing calls at 15 hours long! I
> need to roll another application out and would like to have this problem
> resolved before I do.
>
> Thanks in advance!
> Andy
>
|
| Post Reply
|
|
|
|
|
|
|
|
|
|