Forked from thread: "New NAT66 Draft".

On Mar 10, 2009, James Woodyatt wrote:

My managers don't think we need NAT66.  They think that if we're all
going to be dealing with ubiquitous IPv6 address translation anyway,
then *NAPT66* in our products will be easier to implement and involve
less technical and UI risk for us.  And they're right.  How do I
convince them that NAT66 is worth the extra energy?


James and all -

I wrote an analysis that details why NAT is superior to NAPT.  It looks
at the potential problems that either technique can cause, and evaluates
the cost and completeness of solutions to those problems.

http://users.piuha.net/chvogt/pub/2009/draft-vogt-address-translation-
harmfulness-02.txt

Versions 00 and 01 are in the official IETF repositories, but revision
02 didn't make it before the deadline.  In the analysis, I use the term
"one-to-one address translation" to mean NAT, and the term "many-to-one
address translation" to mean NAPT.

Comments appreciated.

- Christian


_______________________________________________
nat66 mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/nat66

Reply via email to