I think this may be looking at the problem in the wrong way. The one compelling reason to get IPv6, in my opinion, is because it restores the end to end connectivity. IPv6 has enough addresses and ISP can give me my own block that I can assign to various devices in my apartment. The compelling part of IPv6 is that eliminates the need for me to have a NAT. If it weren't for NAT, Ipv4 works well enough, and we don't need IPv6. I would expect that the killer applications for IPv6 will be the ones that take advantage of IPv6's end to end connectivity. If some bozo does try to implement a NAT for IPv6, it will break those applications that make IPv6 compelling.
As an implementor of certain bits of software shipping with a mainstream operating system, I will do my part to make sure that no NAT traversal solutions for various pieces work with IPv6. I've recently been working on IPSec NAT Traversal. The spec notes how the protocol can detect and work around NATs on IPv4 and IPv6. Our implementation intentionally does not support NAT traversal over IPv6. If we can eliminate all of the gross hacks to make software work with NATs in the IPv6 versions, we could create a world in which a NAT can not be deployed because it breaks too much stuff. A vendor could not sell an IPv6 NAT in to such a world.
-josh
On Thursday, March 20, 2003, at 5:05PM, Quality Quorum wrote:
On Thu, 20 Mar 2003, JORDI PALET MARTINEZ wrote:
After the today's decision with site local, is clear to me that we don't
want to have NAT happening again ;-)
We know that the people will do it anyway, but we must do an effort to avoid is as much as possible, and some ideas that could support this are:
IETF is big and strong, market[lace is much bigger and much stronger, in
all previous collisions with market IETF did not fare well.
Note - I am note arguing that NAT will win - my point is that if people want it the resistance is futile.
1) Clearly show the advantages of end-to-end and no NAT model.
2) Have the specs indicating that an IPv6 node (host/router, whatever) MUST NOT support NAT or equivalent mechanisms. Any
interoperability/conformance test must fail if you fail to agree with this specification. This should be a clear sign for the
manufacturers to avoid supporting NATs.
3) Indicate that if someone wants to keep using NAT, should do it with IPv4.
I'm not sure if the rest agree and what is the correct document to say this,
may be as part of the changes for the local-link
deprecation ?
Regards, Jordi
Thanks,
Aleksey
BTW, I prepared a draft which spells out NAT6, it is absolutely trivial thing compaing to NAT4, so if somebody would want it, they would not need any permissions.
***************************** Madrid 2003 Global IPv6 Summit 12-14 May 2003 - Register at: http://www.ipv6-es.com
-------------------------------------------------------------------- IETF IPng Working Group Mailing List IPng Home Page: http://playground.sun.com/ipng FTP archive: ftp://playground.sun.com/pub/ipng Direct all administrative requests to [EMAIL PROTECTED] --------------------------------------------------------------------
-------------------------------------------------------------------- IETF IPng Working Group Mailing List IPng Home Page: http://playground.sun.com/ipng FTP archive: ftp://playground.sun.com/pub/ipng Direct all administrative requests to [EMAIL PROTECTED] --------------------------------------------------------------------
-------------------------------------------------------------------- IETF IPng Working Group Mailing List IPng Home Page: http://playground.sun.com/ipng FTP archive: ftp://playground.sun.com/pub/ipng Direct all administrative requests to [EMAIL PROTECTED] --------------------------------------------------------------------
