Summary:ASTERISK-01298: libpri does not support "SDN Marking"
Reporter:mschaefe (mschaefe)Labels:
Date Opened:2004-03-26 07:00:23.000-0600Date Closed:2011-06-07 14:04:59
Versions:Frequency of
Environment:Attachments:( 0) libpri.patch
( 1) pri_q931.patch
( 2) q931.patch
Description:SDN Marking is required on some PRI's to do outbound dialing.  AT&T uses the Network Specific Facility information element to mark calls as SDN.  This would also support Megacomm using the same ie number with a different NSF code.


I've implemented this in the libpri source code, but I had to make it a mandatory ie.  That would break libpri for other users, so I'm mainly asking how I need to pass flags into libpri for things like these where it is an ISDN option provided by the telco.
Comments:By: Brian West (bkw918) 2004-03-26 17:03:29.000-0600

ifdef it so if people want it they can define that.  Modify the makefile to include such options.  That sounds most logical.  Post a diff and we can all look and see how the best way to do it.

By: mschaefe (mschaefe) 2004-03-26 22:44:04.000-0600

Is there a better way than Makefile to make it configurable on a per-PRI basis?  I am using Asterisk as a switch to front some devices that expect PRI signalling.  I'm not sure I want to make it a mandatory ie because I may want to front simpler PRI devices (e.g. a Dialogic Card) which does not support the feature.  If I add NSF to the list of mandatory ie's, libpri will reject any calls that don't have the ie, right?  I'm going to work on it a bit tomorrow and hopefully submit a patch.

By: mschaefe (mschaefe) 2004-03-31 17:47:54.000-0600

Sorry about the complexity of the patch.  I've been working on both 1306 and 1307 and solved them both.  1307 is due to a switch to Codeset 6 being sent over the PRI.  That ie is ignored, but the next IE is Change Status, and that causes Asterisk to reject the SETUP - not sure why that code exists.  We'll definitely need to discuss this patch.  FYI: I'm pursuing the disclaimer through our legal staff.  Should be no big issues.

By: twisted (twisted) 2004-04-29 09:28:25

any update on this?  disclaimer?  has the code been ifdef'd?

By: mschaefe (mschaefe) 2004-04-30 10:41:45

It's not going to be an easy process to get the code disclaimed.  I can walk you through my final solution so that it's "your" code.  I ended up making it a configuration option in zaptel.conf because it is a feature that is enabled on some PRI's and disabled on other PRI's.  I wanted to support outbound dialing on lineside T1's and didn't want to have to configure them to dial SDN.

There's really no reason to ifdef the code since it's a PRI option and it implements an existing ie that libpri doesn't recognize.

By: Brian West (bkw918) 2004-04-30 12:55:50

Is your company just not wanting to disclaim the fixes? Or do they not understand the whole open source stuff?

By: mschaefe (mschaefe) 2004-05-03 10:04:22

In a large company it's always much easier to say "no" than to say "yes."

By: Brian West (bkw918) 2004-05-05 01:34:09

hehe I understand :)

By: Mark Spencer (markster) 2004-05-21 02:01:37

If we can't get the code disclaimed we mind as well close it out.

By: mschaefe (mschaefe) 2004-05-21 07:36:01

I'm going to implement the code at home on my own time with my own resources based on the AT&T spec and then I can disclaim the code myself.  Please delete the patches, since they are not disclaimed.

By: twisted (twisted) 2004-05-21 09:49:51

Actually, I'm gonna do one better.   Feel free to post back here and re-open this bug once you get things done at home, or, file a new bug when you're ready to impliment this.