(replying to my own message, below)

Adam Roach wrote:

> Vijay K. Gurbani wrote:
>
>> Adam Roach wrote:
>>
>>> In practice, the received parameter is interpreted *almost* exclusively
>>> by the entity that wrote it -- so interoperability shouldn't be as
>>> substantially harmed by differing interpretations than might it first
>>> appear.
>>
>>
>>
>> That depends.  A proxy that inserts a received parameter and *then*
>> forwards the request to the next downstream server may cause some
>> concern to the receiving server.  The specific behavior we have
>> observed is that the next downstream server is unable to parse the
>> request because of the brackets in the topmost Via "received"
>> parameter that were added by the previous upstream entity.
>
>
>
> Why are they parsing anything but the topmost Via header field? It 
> seems a waste of processing time to parse the entire stack of Vias.


To make sure we're not talking past each other: if a proxy inserts a 
"received" parameter and then forwards the request WITHOUT ADDING ITS 
OWN VIA HEADER FIELD on top of the one that it tagged with "received", 
then it's broken. If a proxy is slapping "received" parameters on the 
topmost Via header field that they just added, then it is similarly broken.

Now, the *upstream* server may have some issues in parsing the Via 
header field in responses (and I called this out in my previous 
message), but the contents of the "received" parameter really shouldn't 
be of any concern to them (unless they're using the "received" parameter 
for STUN-like functionality, in which case I think the correct answer is 
"use STUN").

/a
_______________________________________________
Sip-implementors mailing list
[email protected]
https://lists.cs.columbia.edu/cucslists/listinfo/sip-implementors

Reply via email to