[e2e] disintermediation?

[Joe Touch]

Lynne wrote:


> Joe brings up an interesting observation, however. If the receiver of the 

> information presented is unaware of modification of the contents, and the 

> sender of the information is unaware of modification of the contents, and 

> the ISP is the source of the modification, then disintermediation can occur.


Despite efforts to legitimize this by calling it 'transparent proxying' 
or the like, this is clearly a man-in-the-middle attack.

I considering the case where the data is NOT modified, and the header is 
NOT modified, but that the routing depends on the _content_ of the packet.

Right now, the expectation is that packets are sent into the network and 
that policy routing decisions may occur, but only on the header. Peeking 
inside to provide a 'transparent differentiation of service' is the 
issue that doesn't appear to be covered by the former two cases (vanilla 
header-only routing, header or data modification).

Joe