Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.

Messages - looserouting

Pages: [1] 2
Yate bugs / Re: iLBC negotiation
« on: March 20, 2015, 07:40:50 AM »
many UA's are buged and yate is too. ( speaking of ilbc of course  :D )

ilbc_forced seems to be a quick and dirty hack that doesn't always work. Especially beacuse it's global but you are dealing with normal UA's - where you don't need the hack - and with buged UA's.  you only want to set it for particular users otherwise you'll brake compatibility with working UA's.

My approach is to first fix yates behavior. Than you could make a channel-variable out of ilbc_forced to deal with particular UA's.

right know it all seems messed up

Yate bugs / Re: iLBC negotiation
« on: March 20, 2015, 03:08:23 AM »
I think this the following loop in SDPMedia::update that needs change

Code: [Select]
else if (m_formats && !force) {
    // from received list keep only already offered formats
    ObjList* l1 = tmp.split(',',false);
    ObjList* l2 = m_formats.split(',',false);
    for (ObjList* fmt = l1->skipNull(); fmt; ) {
if (l2->find(fmt->get()->toString()))
    fmt = fmt->skipNext();
else {
    fmt = fmt->skipNull();

ilbc20 should not be removed when there is ilbc30 in ObjList l2 and ilbc30 should not be removed when there is ilbc20 in ObjList l2.
For that the else part must be extended to something like:

Code: [Select]
if (  the format is ilbc20 and there is ilbc30 in ObjList l2) {
                    put ilbc30 in fmt instead of ilbc20;
}elseif  ( the format is ilbc30 and there is ilbc20 in ObjList l2)  {
                    fmt = fmt->skipNext();
}else {
    fmt = fmt->skipNull();

I don't know yet which opportunities the Object list class and the string class give me. I am no real programmer and I will nedd some time to fully unterstand the code. and i don't even really know if this is the right approach

also something similar have to be done if only a single format is received.(few line above that code)

comments on this are appreciated

Yate bugs / Re: iLBC negotiation
« on: July 21, 2014, 12:55:29 PM »
I think you don't understand the Problem. Otherwise you wouldn't call it a feature request. The negotiation of ilbc is broken. I try to find some time. Maybe a can write a patch.

Yate bugs / Re: iLBC negotiation
« on: July 18, 2014, 08:33:37 AM »
Existing media must be updated if yate offers mode20 end romote endpoint offers mode30 in SDP. So this function need to be modified to switch ilbc20 and ilbc30 in that case.

There is nothing you can do in [hacks] to make it work correctly.
There is only one option: Disable ilbc20 in [codecs]
The thing is i wanna use ilbc20 where possible.

Edit: i think since for g729 you need a license ilbc20 seems to be the best choice for a narrowband coedec. Other codecs are less common or not as good

Yate bugs / Re: iLBC negotiation
« on: June 05, 2014, 07:24:26 AM »
i can provide examples if you really need them.

I think you're right. The probem is that yate have to switch the format.
Isn't SDPMedia:update the place where this should/could be handled ?

So for now if someone wants to use iLBC with yate he must disable ilbc20 or force mode=30 through ysipchan.cfg.

Yate bugs / iLBC negotiation
« on: June 04, 2014, 07:55:21 AM »
when yate offers iLBC mode=20 and the remote UA anwers with mode=30 then yate must use mode=30. But it doesn't....

AFAIK iLBC mode=20 must only be used if both prefer mode=20.

Yate users hangout place / Re: Is Yate support IPv6?
« on: November 06, 2013, 05:14:09 AM »
i did not recognize that yate5 was released. congratulations !!

Yate users hangout place / Re: Is Yate support IPv6?
« on: October 21, 2013, 02:52:48 AM »
IPv6 is now supported for at least sip and rtp, looking at recent svn commits.

how do you handle such a situation ?

should i send a 480 with the regexroute module ?
But i would like to resolve the problem through the regfile module. this seems to be a better/faster solution for me.

Yate users hangout place / Re: Transcoding and ptime
« on: August 22, 2013, 08:58:58 AM »
thank you for the fix :)

Yate users hangout place / Re: Mapping of Response codes SIP<-->ISUP
« on: August 20, 2013, 04:37:53 AM »
i looked a bit into the source code and it seems to me that no Q.850 cause codes where put into SIP packets.

Yate users hangout place / Re: Mapping of Response codes SIP<-->ISUP
« on: August 07, 2013, 08:31:35 AM »
isn't it a main criteria for a carrier ? For me it's important and I thought for other voip engineers maybe too.

I am using regfile and regexroute for routing.
regfile first then regexroute.
the users are all SIP-users.

i am using regfile for routing to local users and regexroute to call out through my voip provider(configured in accfile).
now if i call a user and he is registered then regfile will route the call.
but if he's not the call will be routed through regexroute.

how can I send an 480 (or user absent for other channels then sip) if the user is offline ?

if i don't load regexroute then yate will send a 480 response. but then i can't call out :)

Yate users hangout place / Re: Mapping of Response codes SIP<-->ISUP
« on: July 30, 2013, 07:37:07 AM »
yate changes a "488 Not Accabtable Here" to "487 Not Accaptable Here" when one SIP Client calls another SIP Client through yate.

Pages: [1] 2