On 2010-10-07 13:57, Fernando Gont wrote:
On 06/10/2010 05:40 p.m., Keith Moore wrote:
It's perfectly reasonable for applications to include IP
addresses and port numbers in their payloads, as this is the only
way that the Internet Architecture defines to allow applications
to make contact with particular processes at particular hosts.
Some might see this as a deficiency in the Internet Architecture,
but that's the best that we have to work with for now.
If anything, the fact that "this is is the only way that the
Internet Architecture defines..." doesn't make it reasonable.
So basically you're arguing to impair the ability of applications to
function, just so that network operators can futz around with
addresses.
No. I'm arguing that you should not blame NATs for broken application
designs, and that you should not assess reasonable-ness based on
existing (and questionable) application designs.
The problem is that the creation of disjoint addressing realms
(due to NAT and to IPv4/IPv6 coexistence) has made distributed
application design almost impossible without kludges.
See draft-carpenter-referral-ps-01
Brian
_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
https://www.ietf.org/mailman/listinfo/ietf