Re: ARe: Abstract Model contribution for Intermediairies (repost)

According to your model, the initial path specified by the sender can be
modified by
intermediaries.  Based on this idea, intermediaries should be able to
modify the response path,
that is, the reverse path in your term.  In my example, the sender side
gateway "modifies"
the reverse path so as to forward the response message to the intermeiary
in Intranet.
I think this is a natural extention of your idea, and organizes a symmetric
structure.
Make sense?

>Yuhichi Nakamura wrote:
>> [...]
>> For the first comment, I would restate my question: if there is an
>> intermediary BEFORE the gateway, MUST it be recored in the path record?
The
>> company who has the intial sender, the gateway, and the added
intermediary
>> would not like to give the address of the intermediary because it is
located
>> in their intranet.
>> [...]
>Ah, I see what you mean! Yes, you are right, this is an argument for NOT
>recording the whole route. Does that mean that, on the way back, the
message
>will NOT go through the intermediary at all?
>Jean-Jacques.
>


Yuhichi Nakamura
IBM Research, Tokyo Research Laboratory
Tel: +81-46-215-4668
FAX: +81-46-273-7428

Received on Thursday, 15 March 2001 07:31:41 UTC