On Mon, 2009-11-02 at 14:48 -0500, Dale Worley wrote:
> On Mon, 2009-11-02 at 05:46 -0800, Scott Lawrence wrote:
> > On Sun, 2009-11-01 at 17:32 -0500, Dale Worley wrote:
> > > On Sun, 2009-11-01 at 13:08 -0800, Scott Lawrence wrote:
> > > > Why do you conclude that sipXbridge modifies the contact for the benefit
> > > > of the proxy?   
> > > 
> > > A couple of days ago I talked to Ranga, who mentioned that the proxy
> > > does not handle maddr parameters correctly, and I have some vague memory
> > > that he said that sipXbridge removed them.  In any case, we have to
> > > worry about the situation, since if sipXbridges does *not* remove the
> > > maddr parameter, the proxy *will* handle the request incorrectly.
> > 
> > I don't understand... if there was a bug in the proxy not handling maddr
> > correctly, why wouldn't we _fix_ it instead of doing a workaround in
> > sipXbridge that is likely to break something else?
> 
> I didn't write it, I am just observing.
> 
> > Looking at the trace in 6909, it's clear that sipXbridge is breaking the
> > routing between frames 52 and 53 - it strips the maddr but just drops
> > the information.... 
> > 
> > It might have worked if the bridge had added the maddr value to the
> > record-route to preserve the routing, but that's probably not the right
> > thing to do.
> 
> More interestingly, what is the right thing to do to fix XX-6909?

1. Change sipXbridge so that it doesn't discard maddr.
2. Fix whatever that breaks.

_______________________________________________
sipx-dev mailing list [email protected]
List Archive: http://list.sipfoundry.org/archive/sipx-dev
Unsubscribe: http://list.sipfoundry.org/mailman/listinfo/sipx-dev
sipXecs IP PBX -- http://www.sipfoundry.org/

Reply via email to