On Monday 9 January 2006 11:08, you wrote:
>
> I partially understand, but for me this will work, so I'll keep the patch
> for myself and apply whenever updating to a newer patch level! ;-)
I may be adding a workaround based on a filter which specifies which
extensions do support overlap dialing so, it won't last for too much. But I
have to work on the EC, first :)
> (Partially because I do not 'quite' grasp what you mean by "The problem is
> that when the overlap dialing fall in an extension which does support
> overlap dialing I have to start the PBX immediately and pass further
> digits as DTMF frames")
Suppose you have a box with two ISDN interfaces configured for passthrough.
One in TE mode connected to the telco, one in NT mode connected to an ISDN
phone.
When you hang off the phone it sends an empty SETUP, asterisk starts the PBX
on the s,1 extensions which in turn has VISDNOverlapDial() application that
also provide the dialtone.
Then I start dialing the digits one by one.
Suppose you have an entry like this in your dialplan and you want to dial
01234:
exten => _0.,1,Dial(VISDN/visdn.telco/${EXTEN})
After you digit '0', I have to start Dial() which will send a SETUP with '0'
towards visdn.telco interface and start overlap dialing on it.
All further digits (coming as INFORMATION messages) will be passed between
channels as DTMF frames and recoverted to INFORMATION messages again.
When the dialed channel receives a PROCEEDING, I pass a PROCEEDING frame back
and a PROCEEDING message on the dialing channel.
This works well and makes ISDN-ISDN calls really transparent.
Now suppose you have another extension and the only valid SIP address is 2987:
exten => _2.,1,Dial(SIP/${EXTEN})
You start dialing 2 and Dial(SIP/2) is run, which does not correspond to any
valid address and Dial exit with NOT FOUND...
> Another thing I am looking at right now is that the chan_visdn doesn't
> recognise the DTMF after the call is established, which means I cannot use
> functions like ATXFER, XFER, AUTOMON, etc., because these rely on
> key-presses (resp. *2, # and *1 by default), or even terminate the echo
> feedback test (*43) by pressing #...
Yes, this is known, I just have to enable the DTMF detecton integrated into *
and maybe use the hardware DTMF detector in some hardware.
There is a TODO entry already, I'll work on it just after the EC.
Bye,
--
Daniele Orlandi
_______________________________________________
Visdn-hackers mailing list
[email protected]
https://mailman.uli.it/mailman/listinfo/visdn-hackers
