Hi!
> Does * have the capability to screen calls? IOW, if someone calls in from
> outside (ie. not a local extension), can * ask the calling party to state
> their name, record it, ring the recipient, play the caller's name for the
> recipient, then give the recipient the choice of answering or forcing the
> call to voice mail?
Hm, just an idea:
- enable recording (see application "monitor")
- answer the incoming call
- playback an annoncement and request the user to speak the name,
followed by some digit to end the recording
- on key press issue a StopMonitor (you'll need an AGI script for this)
- transfer the person to call parking (or into a conference room with
MOH)
- call the target and use playback the file recorded with monitor
- now announce the call park number (where would you get that from?), or
allow the target to administrate (how?) the conference room
- create an extension that triggers another AGI script that hangs up on
the caller if the target doesn't want to talk, or let the target use
astman/gastman
Sounds a bit complicated, I admit.
Cheers, Philipp
Or, you could apply my patch, that I've been upgrading on the asterisk bug site.
Check http://bugs.digium.com/bug_view_page.php?bug_id=0000752
See if the drawbacks outweigh the features. I'm working on it right now, have a few
things I need to conquer, but it does work if you play music on hold during the dial,
and everyone stays away from the # key.
As to what the stock Asterisk supplies for call screening ("Privacy"):
Already built into asterisk is the "privacy" app, which forces the caller to enter a 10 digit
phone number, if they are calling anonymously. Also available are some C routines to
check the privacy database, and the "privacy" family in the database. Entries in the database
are by CID, and the value can be ALLOW, DENY, KILL, and TORTURE. Allow means to
let a caller from a CID straight thru. DENY means to just connect them straight to email.
KILL means to play them a short "go away and don't come back" type message, and TORTURE
means to send them to a context that will make them suffer. I've invented on that hoped to
do something along that line.
And finally, the Dial application allows you the P option, which just basically has code to collect
the optional database name (if you specify P(dbname)). Here is where I've been playing, and if
you check the bug number I mentioned above, you'll get all the gory details, but basically it does
as you hoped.
As a side note, in the area of Caller/Callee anouncing:
A side affect of collecting little "intros" via the privacy option, is that you can use them
to play the intros over a PA.
I found that if I take the last couple month's logs, and compile a list of folks that have
called in, and generate (record into a .gsm file) "intros" for them, I can use these intros to
play their name over a speaker when calls first come in. I play the voicemail greeting wav file
when they select an extension. Around here, I have 6 kids on one phone, and they
can't stand the suspense of finding out who a call is for! Pre-recording such intros helps
reduce the frustration of people calling in, and supplies you with a high hit rate right from
the start. After that, new callers add their contributions to the list.
murf
signature.asc
Description: This is a digitally signed message part
