asterisk users - Sep 2004 - Asterisk 1.0 & Zaptel 1.0 -- False Hangup Disaster

I was really looking forward to Asterisk 1.0 et al, but it is a major
disappointment. I have never experienced any Asterisk release that was
interacting with Digium hardware so unreliably.

Asterisk hangs up on every outgoing PSTN call (via Zaptel) as soon as
the call is being picked up at the other end.

I have tried various X100P (original Digium) cards, various phone
lines and just about every possible combination of parameters in
zaptel.conf (from every feature off to most features on) but without
any success.

The same hardware was working fine with CVS from about a month ago.


The problem of false hangups really needs to be fixed. A false hangup
is NEVER EVER acceptable in an office environment. On the other hand,
a call that doesn't hangup even if the remote party has already hung
up is ALWAYS acceptable. Therefore, if the software is not capable of
detecting hangups properly, then why not provide a setting to disable
any and all hangup decisions now made by software and let the human
user decide instead. There should be a setting "hangup=local-only"
that would have the effect that no channel will ever be hung up unless
the (non-Zap) local party has hungup.

As things stand now, we won't be able to deploy this 1.0 release if
Zaptel is required. What a pity.

rgds
benjk

-- 
Sunrise Telephone Systems, 9F Shibuya Daikyo Bldg., 1-13-5 Shibuya,
Tokyo, Japan.

NB: Spam filters in place. Messages unrelated to the * mailing lists
may get trashed.

On Saturday 25 September 2004 08:03, Benjamin on Asterisk Mailing Lists
wrote:
> Asterisk hangs up on every outgoing PSTN call (via Zaptel) as soon as
> the call is being picked up at the other end.
Disable callprogress and/or busydetect.
> I have tried various X100P (original Digium) cards, various phone
> lines and just about every possible combination of parameters in
> zaptel.conf (from every feature off to most features on) but without
> any success.
Where are you located?  Is your phone company perhaps giving you call 
completion supervision?  Your post is quite short on details.
> The problem of false hangups really needs to be fixed. A false hangup
> is NEVER EVER acceptable in an office environment. On the other hand,
> a call that doesn't hangup even if the remote party has already hung
> up is ALWAYS acceptable. Therefore, if the software is not capable of
I disagree wholeheartedly; call startup and termination must both be reliable.  
Having 5 minutes of busy tones on VM is not acceptable.

So let's get to the root of your problem.
> user decide instead. There should be a setting
"hangup=local-only"
> that would have the effect that no channel will ever be hung up unless
> the (non-Zap) local party has hungup.
You must be new here.
> As things stand now, we won't be able to deploy this 1.0 release if
> Zaptel is required. What a pity.
With that attitude, we won't even miss you.

-A.

On Sat, 25 Sep 2004 21:03:31 +0900, Benjamin on Asterisk Mailing Lists
<benjk.on.asterisk.ml@gmail.com> wrote:
> 
> The problem of false hangups really needs to be fixed. A false hangup
> is NEVER EVER acceptable in an office environment. On the other hand,
> a call that doesn't hangup even if the remote party has already hung
> up is ALWAYS acceptable. Therefore, if the software is not capable of
> detecting hangups properly, then why not provide a setting to disable
> any and all hangup decisions now made by software and let the human
> user decide instead. There should be a setting
"hangup=local-only"
> that would have the effect that no channel will ever be hung up unless
> the (non-Zap) local party has hungup.
I think that letting the non-Zap handle the hangup is a very good
idea. If the non-Zap originated the call, than it has the right to
terminate it. If it didn't, it will eventually perceive that the
caller hung up, and won't spend the day on a dead phone...

It is better than the call being disconnected in the middle of an
important discussion, and it may create the impression that the other
person slammed the phone on you if you were arguing or something like
that...

The only problem is incoming calls to IVR, VM, and such.

It should be allowed to specify if busy detect is enabled only for
zap-originated, zap-terminated, or both kinds of calls.

That's my opinion... :)
Marconi.