Re: Operations requirement: no hop-by-hop

At 11:22 AM +0100 2/15/04, Hadmut Danisch wrote:
>On Fri, Feb 13, 2004 at 10:54:37AM -0800, Paul Hoffman / IMC wrote:
>>
>>  This may or may not be controversial, but I propose that a
>>  requirement for sane operations is that mail-ng be single hop.
>>  Multi-hop was a good design 20 years ago when many entities were not
>>  online all the time, but it has also proven to be a huge hurdle in
>>  extending SMTP.
>>
>>  From an operational standpoint, a single-hop design makes many other
>>  parts of the design much cleaner. It is also easier to explain to
>>  users.
>
>
>
>I'd propose to split the transport (and discussion) into two
>parts:
>
>The traffic between different domains (or whatever we will have
>in future), and the intra-domain traffic.
>
>
>For the traffic between domains, single-hop is considerable.
>But for the traffic within the domain (i.e. what is
>MX to mailbox today), you can't assume every mailbox to be
>online all the time.

Fully agree. I didn't get to that point until the end of the message, 
but I can see from other responses that people missed it. Let me try 
again:

The mail-ng transport protocol should be designed to only know about 
a single hop, from sender to recipient. The sender may have multiple 
hops to a designated transport agent, and the receiving transport 
agent may have multiple hops to a mailstore, but the transport 
protocol should not make any accommodation for these additional hops 
other than to allow them to add metadata to the transit stream. These 
requirements are to make the protocol more sensible and to limit the 
responsibility of operators to just what they control.

--Paul Hoffman, Director
--Internet Mail Consortium