17 Oct
2005
17 Oct
'05
9:36 a.m.
Many people pick this up and twist it into ~the network has to be application agnostic~ and then use this against NATs or firewalls, which is simply a misuse of the principle.
Personally, I think that NAT's interference with the communication between hosts is similar to the way in which error-detection and retransmission interfere with realtime voice communication, as described in Saltzer's end-to-end paper: http://web.mit.edu/Saltzer/www/publications/endtoend/endtoend.txt It seems that the end-to-end principle is more of a metaphor for how to look at the design problem rather than a hard and fast rule. http://www.postel.org/pipermail/end2end-interest/2002-March/001848.html --Michael Dillon