[kwlug-disc] OT: Roger's phone services

[Andrew Kohlsmith (mailing lists account)]

On Monday, July 19, 2010 01:13:01 pm John Van Ostrand wrote:
> > I was talking more of the setup where the Nortel system is being
> > augmented
> > with VOIP; thus you'd have FXO ATA ports going to the telco and FXS
> > ATA ports
> > going to the Nortel.

> Yup, we're in sync. Incoming call hits the ATA which gets the CID from the
> SIP header. It then rings the Nortel, which does it's normal thing and

Not if it's a call from POTS; that's what I was referring to.  There are both 
FXS (to Nortel) and FXO (from telco) ports; incoming SIP calls have no delay, 
you are right. But incoming landline calls will have this delay.

> waits for two rings. That part is normal and shouldn't cause any
> additional delay. If the call then goes to a Nortel handset, it's all
> normal. It's only when it goes to a Nortel ATA, then SIP ATA that a
> additional two rings occur. This would be after a voice menu, so although
> the caller is inconvenienced there isn't a very long ring happening. Only
> if the caller was directed to an ATA phone directly, without a voice menu
> would the caller hear 4 rings too many.

That's the other location for extraneous rings.  Isn't analog telephony fun? 
:-/

> Using DID features, the call goes directly to the extension. (yeah, yeah, 4
> ring delay.)

The Meridian systems I had the pleasure of working with had a fixed (and 
absurdly low) number of DIDs they'd allow to ring an internal extension, so we 
had to use the macro+IVR method. Unfortunately that doesn't work well if the 
person's on the phone, and bridging the two audio paths (asterisk ring + 
nortel's "ringing and extension" ring) caused audio blips that kind of took 
away from the "magic."

-A.