IAXmodem

Krzysztof Drewicz wrote:
> Hello,
>
> i wonder if there is a catch that i'm not aware of, what i do is a lot
> of outbound faxing.
>
> I get on asterisk side: ISDN cause code like "1" - no such number or
> "27" - circuit busy etc..  (full list for example is here:
> http://cnes.com/causecodes.html)
>
> So, i whould like to get, the asterisk to pass cause code to iaxmodem
> (not so hard as i thought) but, then, how iaxmodem should reply this
> to hylafax?
> Some thing like "+FHNG:2" or similar ? (ModemAnswerCmd at a
> hylafax-config man page)
>
> Any hints?

IAXmodem is a Class 1 modem and +FHNG: responses are for Class 2, and 
furthermore +FHNG already has a defined set of responses, etc, so if we 
were dealing with a Class 2 modem (and we're not), then you'd need to be 
careful to avoid colliding with the pre-defined responses.

There really is no place to add a similar type of debugging message in a 
Class 1 modem if we limit ourselves to the T.31 spec.  However, since 
v5.1.6 HylaFAX+ has been able to cope with +FDB (debug) messages which a 
select few modems use when running in debug mode.  So, I would suggest 
that the right approach would be to use an iaxmodem config option to 
turn-on +FDB debug messages.  So it would be something like +FDB:xxx 
where xxx will refer to whatever debugging information iaxmodem wants to 
pass-through.  You'll be limited to using this feature with HylaFAX+, 
however, until other fax applications adopt the +FDB responses.

How, exactly, are the ISDN cause-codes passed through to IAXmodem, however?

Thanks,

Lee.

2011/10/7 Lee Howard <faxguy@...>
>
>
> IAXmodem is a Class 1 modem and +FHNG: responses are for Class 2, and furthermore +FHNG already has a defined set of responses, etc, so if we were dealing with a Class 2 modem (and we're not), then you'd need to be careful to avoid colliding with the pre-defined responses.


Yeap. Thanks for pointing it down. So this is not to be used.


>
> There really is no place to add a similar type of debugging message in a Class 1 modem if we limit ourselves to the T.31 spec.  However, since v5.1.6 HylaFAX+ has been able to cope with +FDB (debug) messages which a select few modems use when running in debug mode.  So, I would suggest that the right approach would be to use an iaxmodem config option to turn-on +FDB debug messages.  So it would be something like +FDB:xxx where xxx will refer to whatever debugging information iaxmodem wants to pass-through.  You'll be limited to using this feature with HylaFAX+, however, until other fax applications adopt the +FDB responses.
>
> How, exactly, are the ISDN cause-codes passed through to IAXmodem, however?
>

As per RFC5456 (The IAX2 spec)  in 6.2.5. HANGUP Request Message there
are a two optional sections Cause and Cause Code (more info in this
rfc under Section 8.6.21 & 8.6.33 )
It is somehow only limited to a ITU-T range of 1..127 (not for example
a SIP error codes), and one should be aware that you could have answer
from not a ISDN trunk but from a SIP provider or ISUP,  and this puts
some reasons 'in one bag' just for translating:
ISUP -> SIP
4 Call forward; line busy               181 Call is being forwarded
5 Call forward; no reply                181 Call is being forwarded
6 Call forward; unconditional           181 Call is being forwarded
or:
34 no circuit available                 503 Service unavailable
38 network out of order                 503 Service unavailable
41 temporary failure                    503 Service unavailable
42 switching equipment congestion       503 Service unavailable
47 resource unavailable                 503 Service unavailable

But, in real world, I would love to have a +FDB in iaxmodem just
reporting ISDN cause, and well, i use hylafax+, so no point in saying
it's a limit, it is a "+" after all?
When i will find some time i think i will implement this in +FDB
message, is there any guide-line or format to use?

Thanks for your reply.

Krzysztof Drewicz wrote:
> 2011/10/7 Lee Howard <faxguy@...>
>   
>
>> There really is no place to add a similar type of debugging message in a Class 1 modem if we limit ourselves to the T.31 spec.  However, since v5.1.6 HylaFAX+ has been able to cope with +FDB (debug) messages which a select few modems use when running in debug mode.  So, I would suggest that the right approach would be to use an iaxmodem config option to turn-on +FDB debug messages.  So it would be something like +FDB:xxx where xxx will refer to whatever debugging information iaxmodem wants to pass-through.  You'll be limited to using this feature with HylaFAX+, however, until other fax applications adopt the +FDB responses.
>>
>> How, exactly, are the ISDN cause-codes passed through to IAXmodem, however?
>>
>>     
>
> As per RFC5456 (The IAX2 spec)  in 6.2.5. HANGUP Request Message there
> are a two optional sections Cause and Cause Code (more info in this
> rfc under Section 8.6.21 & 8.6.33 )
>   

You can make this change to your iaxmodem.c and then watch the iaxmodem 
output to experiment with this.

--- iaxmodem-1.3.0/iaxmodem.c   2010-03-01 10:12:48.000000000 -0800
+++ iaxmodem/iaxmodem.c 2011-10-07 10:03:45.494794888 -0700
@@ -1288,7 +1288,7 @@
                        if (gothup) sighandler(SIGHUP);
                        break;
                    case IAX_EVENT_HANGUP:
-                       printlog(LOG_INFO, "Remote hangup.\n");
+                       printlog(LOG_INFO, "Remote hangup. Cause: \"%s\" 
Code: %u\n", iaxevent->ies.cause ? iaxevent->ies.cause : "", 
iaxevent->ies.causecode);
                        if (modemstate != MODEM_ONHOOK) {
                            if (modemstate == MODEM_CALLING)
                                t31_call_event(&t31_state, 
AT_CALL_EVENT_BUSY);

> But, in real world, I would love to have a +FDB in iaxmodem just
> reporting ISDN cause, and well, i use hylafax+, so no point in saying
> it's a limit, it is a "+" after all?
> When i will find some time i think i will implement this in +FDB
> message, is there any guide-line or format to use?

To get the +FDB: message requires a bit more work.  The T.31 modem in 
spandsp will require some work to get it to report the message, and then 
some kind of t31_call_event added to the IAX_EVENT_HANGUP case.  I could 
do this myself, of course, but all of my free-time is consumed right now 
for at least the next month.  If you or someone else does the work then 
I'm happy to review it and include it and pass the bits on to Steve 
Underwood for inclusion in spandsp.

Be aware that so far in the very few tests I ran with the change 
mentioned above I haven't seen any non-null "cause" values come 
through.  So this may represent a bug in libiax2 or that Asterisk 1.4 
isn't populating it in the first place... both of which would need yet 
more work to resolve.

I would recommend a T.31 response formatting of something like: 
'+FDB:H%u:%s' where "H" stands for "hangup" and %u is the causecode and 
%s is the cause.

Thanks,

Lee.