Re: [FYI] billing identifier/information proxying

On Wednesday 29 October 2003 16:08, Arne K. Haaje wrote:
> tirsdag 28. oktober 2003, 14:20, skrev Alexander Malysh:
> > Hi,
> >
> > I'm -1 for change of shared.c... sorry but it will not work for at least
> > smpp module...
>
> Why?

because for smpp module service_type should be sent for every sms not not only 
for the first part, I believe for emi is it equal ...

>
> Now that Kannel has support for billing, there should be a policy of how to
> handle this problem of billing mulitpart messages. As implementations may
> vary across operator, could perhaps a configuration parameter solve it?
> Something like bill-first=yes|now. If bill-first=yes then my patch or
> something similar could be used.

heh, yet another config option :( but indeed we should handle it...
what other developers think about this?

>
> > CIMD2 patch look good, but see below ... Unfortunately we don't have any
> > cimd connections there so can't test it :(
> >
> > @@ -2034,11 +2046,14 @@
> >      default:
> >          code = 0;
> >      }
> > +	if (st_ecode > 0)
> > +		code = st_ecode;
> >
> > I believe that this will not work?! here code is a kannel's dlr-mask, see
> > sms.h or are kannel's dlr-mask and st_ecode equal?
>
> It does work for us. We now receive the errorcode when it greater than
> zero, otherwise we receive Kannels usual status code.

I'm sure it works for you. But maybe dlr_find function awaits another value?! 
that doesn't match this st_ecode. This is due to actions defined for dlr 
core. e.g. when code == DLR_BUFFERED then messages will not be removed from 
storage just status flag update will be done, when code == DLR_SUCCESS or 
code == DLR_FAIL then message will be removed from storage and so on. Please 
check it...

>
> >      if(code)
> >      	msg = dlr_find(conn->name, timestamp, destination, code);
> >      else
> >          msg = NULL;
> >
> >
> > On Tuesday 28 October 2003 13:40, Arne K. Haaje wrote:
> >
> > [snip]
> >
> > > Attached is a patch for smsc_cimd2.c that set field 64 Tariff Class to
> > > the value of binfo if binfo is set.
> > >
> > > I have also updated smsc_cimd2.c to set field 55 Reply Path to enabled
> > > if rpi is set.
> > >
> > > These two changes should make the CIMD2 module more up to date, but my
> > > final change to smsc_cimd2.c may be more controversial.
> > >
> > > There is an integer field in the CIMD2 protocol called 62 Status Error
> > > Code that can provide more detail about why a message was not
> > > delivered. When doing MT billing it is necessesary to know if a message
> > > failed because it belonged to an other operator (ported), or if a
> > > prepaid card had no balance.
> > >
> > > I have altered smsc_cimd2.c to send this error code instead of the
> > > usual if it is set. If you feel this is not the way to do it then feel
> > > free to rip it out if you apply the rest of the patch.
> > >
> > > I have also made a change to shared.c that may be a bit controversial.
> > > For multipart messages like picturemessages, I have altered shared.c to
> > > only set binfo for the first part. This is to avoid double of triple
> > > charging the customer! The CIMD2 protocol says nothing about this
> > > issue, but the two operators we use CIMD2 with, want it that way.
> > > However, it may perhaps not work with other modules.
> > >
> > > We run this patched code against two operator, and have had no problem
> > > with it.
> > >
> > > Please discuss and vote.

-- 
Best regards / Mit besten Grüßen aus Düsseldorf

Dipl.-Ing.
Alexander Malysh
___________________________________________

Centrium GmbH
Vogelsanger Weg 80
40470 Düsseldorf

Fon: +49 (0211) 74 84 51 80
Fax: +49 (0211) 277 49 109

email: [email protected]
web: www.centrium.de
msn: [email protected]
icq: 98063111
___________________________________________

Please avoid sending me Word, Excel or PowerPoint attachments.
See http://www.fsf.org/philosophy/no-word-attachments.html