Re: NATmakes a network a host -- must every process have an IP address?

> Sounds like you are tripping across the end-to-end myth.  Whenever we design
> a protocol that believes there is a direct interaction between one end-point
> and the other, we later have to deal with the presence of intermediation.
> Caches.  Firewalls.  Whatever.

another view is that these intermediaries are a fad, an idea that people are only 
starting to realize is a bad one.

another view is that these intermediaries are a sign of a fundamental deficiency
in the architecture - in particular the inability of the network (under the current
architecture) to provide mechanisms that would allow for intelligent traffic
filtering to provide security in depth.

another view is that these intermediaries are mostly due to poor/insecure 
implementation at the hosts.  

however I do not believe that store-and-forward is appropriate for all or even
most applications.  I'm not even sure it's appropriate for email.  The failure to 
directly connect source to destination is one of the principal factors behind
the poor reliability of email.


Received on Friday, 6 December 2002 08:38:29 UTC