[e2e] What should e2e protocols know about lower layers?

[John Wroclawski]

Joe,

At 10:08 AM -0700 10/10/01, Joe Touch wrote:
>
>I guess I'm the one on the other side of the debate.
>
>IPv4 certainly has a notion of local - the subnets to which you are 
>directly connected.

Indeed. But this "local" is from the perspective of things the IP 
layer concerns itself with - routing, addressing, etc.

>
>While I appreciate the use of congestion control throughout, I still 
>maintain that there are times when the network layer in particular 
>needs to distinguish between local and nonlocal. I see no reason why 
>the transport layer shouldn't be afforded the same opportunity.
>

Whether or not there's value in allowing the transport layer to 
distinguish between local and non-local, the _definitions_ of local 
in the two cases aren't the same. As you mention above, the network 
layer definition of local is "directly connected subnet" - 
appropriate for IP-layer things, but completely unrelated to any 
useful definition of "local" from the perspective of transport-layer 
concerns such as congestion management.

If you want to defend the transport having a notion of local vs 
non-local, I think you first have to say what "local" vs "non-local" 
means from the perspective of a transport layer endpoint, and then 
show that you can accurately decide whether a remote endpoint is 
local or not. The notion that "transport-layer-local" is congruent to 
"same IP subnet" just won't hunt.

Cheers,
John