On Tue, May 31, 2011 at 11:54 AM, Iñaki Baz Castillo <[email protected]> wrote:

> 2011/5/31 Brez Borland <[email protected]>:
> > Looking at the issue at hand, Iñaki brings me this concern. What if
> request
> > is injected with the number of Route headers pointing to the local
> > elemen(s)t? How to validate that? rfc3261 is talking about loop detection
> in
> > regards to the Via headers, not Route.
>
> That's the point. I don't want my proxy doing useless loops just
> because an ugly client has included more than N pre-loaded Routes
> pointing to the proxy itself.
>
> In my proxy, I've implemented that the loose-route mechanism removes
> all the Route headers pointing to the proxy (from top-to-bottom) until
> it finds a Route not pointing to the proxy. In my tests it works and
> breaks nothing.
>
>
That is a plausible approach indeed. I would love to see somebody to
contradict that by pointing to the document mandating the opposite.
Otherwise, IMHO that is a clear 'hole' in the SIP specification that's been
overlooked.


Regards,


Brez




> Regards.
>
>
> --
> Iñaki Baz Castillo
> <[email protected]>
>
_______________________________________________
Sip-implementors mailing list
[email protected]
https://lists.cs.columbia.edu/cucslists/listinfo/sip-implementors

Reply via email to