Hi Scott,
|NAT translates from a local locator to a global locator ("RLOC" in the
|terminology you want us to adopt).
Actually, I'm not pushing a given terminology, just documenting what seems
to have evolved.
|You say
|
|> It's not unreasonable to have a name that has both
|> identification semantics and local locator scope.
|
|That's true, and what matters in the case of NAT is the "local locator"
|semantics. NAT is not treating the address field it translates as an
|endpoint identifier, but as an address. That's why I say NAT
|doesn't do
|what you said it does.
Fair point. You're right, the text in 3.1.2 is poor. Let's take a pass at
fixing it:
Translation solutions are characterized by converting an original local
source name to a global name and then from the global name back to a final
local destination name. The global names are assigned in an aggregatable
fashion to provide scalability.
Work for you?
|> |> 3.3.1. Strategy A
|
|If we end up using this classification scheme in our recommendation to
|the IETF, one thing we need to do is name these strategy
|buckets in ways
|that are easier to remember than "A", "B", etc.
Actually, I like it. It's completely neutral, it avoids all of the
arguments about natural language naming, and it clear and succinct.
Tony
_______________________________________________
rrg mailing list
[email protected]
http://www.irtf.org/mailman/listinfo/rrg
