[Asterisk-Users] Delayed ringing on some SIP phones

[Rich Adamson]

> > Wouldn't something like this work for you?
> > [incoming-bri-one]
> > exten => s,1,SetCIDName(Company 1)
> > exten => s,2,Dial(SIP/200&SIP/201&etc.,15) ; comapny 1's phones
> > exten => s,3,Dial(SIP/200&SIP/201&SIP/300&SIP/301&etc.,15) ; 
> > company 1's & 2's phones
> > exten => s,4,Voicemail(su200)
> > [incoming-bri-two]
> > exten => s,1,SetCIDName(Company 2)
> > exten => s,2,Dial(SIP/300&SIP/301&etc.,15) ; company 2's phones
> > exten => s,3,Dial(SIP/300&SIP/301&SIP/200&SIP/201&etc.,15) ; 
> > comapny 2's & 1's phones
> > exten => s,4,Voicemail(su300)
> 
> Thanks for the replies folks.
> 
> My concern is that the SIP phones in question (GXP-2000s) tend to take a
> second or two to realise they're no longer ringing. If phones are ringing
> from the first dial statement and still think they're ringing when the
> second dial statement is executed, they will all report busy to asterisk and
> not ring at all.
> 
> I suppose I could insert a Wait(2) or something like that between the two
> dial statements, but I can see it causing problems with users picking the
> phone up and finding nothing on the end, then when the second dial kicks in,
> their phone reports busy because it's off-hook. What I'm really after is a
> method of starting the second dial on company 2's phones without
> interrupting the dial on company 1's phones.

If one thinks about how sip phones are actually caused to ring (eg, sending
a sip packet with Ring in it), and think about the ring timing sequence
implemented within the phone itself, then one should be able to pick a
timeout value (probably not 15 seconds) to minimize the above probability.

Example... in the US the ring cycle is approximately seven seconds. If the
timeout is set to 15 seconds as opposed to 12 seconds, the probability of
picking up the phone during that unwanted period is different.

Pure guess on my part given the internal speed of asterisk, but I'd suggest
that time interval from leaving one dial statement and moving into the
seocnd is very likely measured in milliseconds. That would imply the user
would have to pick up the phone at just the exact moment covered by that
duration for the unwanted condition to occur. Since the phone system (and
people) are not very busy (one pstn line each), this would appear to be
something that technical folks would argu forever with no real business-
world impact. Even if it did impact the user, call pickup would address it.

Am I really all that far off base with that thought process?