[e2e] source code release

[Yu-Shun Wang]

Hi, Zoltan,

Please see comments inline.

Zoltan Turanyi wrote:
> On Sat, 1 Mar 2003, Christian Huitema wrote:
> 
> 
>>How is 4+4 in any way easier to deploy than 6to4?
> 
> 
> Christian,
> 
> 6to4 is a mechanism to enable connectivity from a local site to remote
> v6 destinations over a v4 cloud. If the local hosts need connectivity to
> remote v4 destinations as well, either dual IPv4/IPv6 routing inside the
> 6to4 site is needed or a IPv6-IPv4 protocol translator. None of these are
> required with 4+4. Existing v4 routers can remain and operate as

I have a different view regarding the question.

Once we have 6to4, you can use IPv6 staight on. Why do we need
6to4 and then worry about v4 connections (over 6to4)? The point is
not whether v4 is easier in 6to4 scheme or 4+4, it's IPv6 vs. 4+4
at this point.

> already configured. The extension in the NAT is simple, stateless and 
> nothing as ugly as protocol or even address translation. (Of course, 
> the address translation function of the "extended NAT" will go away as 
> transition progresses. All that remains is simply a router that 
> understands 4+4 packet format.)

"a router that understands 4+4 packet format" vs. "a native v6 router"

It's a tall order to convince people the former is better than later,
IMHO.

> In addition, it is not necessary to further reconfigure
> routing with 4+4 to *remove* 4+4. A lagre part of RFC3056 discusses
> configuration options and routing rules to co-exist (and eventually
> transit) between 6to4 and native v6 routing. These are not necessary with
> 4+4. Existing routing policy can be kept and since 4+4 is not a transition
> mechanism to something else, you do not need to remove it in the end.

Yes, as a transition mechanism, 6to4 is not a pretty one. But if
4+4 is positioned to be a permanent replacemant of IPv4+NAT, then
the practical question really is whether 4+4 is better than IPv6.
(Is it?)

Regards,

yushun.

> Zoltan


-- 
____________________________________________________________________________
Yu-Shun Wang <yushunwa at isi.edu>               Information Sciences Institute
                                            University of Southern California