2008/4/30 Matt Krokosz <[EMAIL PROTECTED]>:
> I'm trying to override the Via header used for an outbound Invite. There
> appear to be provisions to allow me to specify the Via at the application
> level by doing the following…
> nua_invite(nh, NTATAG_USER_VIA(1),
> SIPTAG_VIA_STR("SIP/2.0/UDP 1.2.3.4"),
> (other stuff)
> TAG_END());
> However, the stack does not seem to take my Via, or at least, my host
> portion of the Via. After looking at the code, it appears that the
> following piece of code in function "outgoing_insert_via()" of "nta.c" is
> causing the problem and the comment looks suspicious as well.
> /* XXX - should we do this? */
> if (via->v_host != v->v_host && str0cmp(via->v_host, v->v_host))
> clear = 1, v->v_host = via->v_host;
> If I comment out the clearing of the host in this scenario, then everything
> works as expected. It appears that this piece of code is not allowing the
> application to specify the host portion of the via, even when using
> NTATAG_USER_VIA(1) with an application specified Via.
> Is this a "bug" in the stack?
Looks like a bug to me. Host is not used for anything very useful,
port is used when returning the response over UDP.
I'll push a patch shortly to darcs repo at sofia-sip.org.
--
Pekka.Pessi mail at nokia.com
-------------------------------------------------------------------------
This SF.net email is sponsored by the 2008 JavaOne(SM) Conference
Don't miss this year's exciting event. There's still time to save $100.
Use priority code J8TL2D2.
http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone
_______________________________________________
Sofia-sip-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/sofia-sip-devel
